(Lecture Notes in Computer Science 10058) Marinos Ioannides, Eleanor Fink, Antonia Moropoulou, Monika Hagedorn-Saupe, Antonella Fresa, Gunnar Liestøl, Vlatka Rajcic, Pierre Grussenmeyer (eds.)-Digital.pdf

Marinos Ioannides · Eleanor Fink
Antonia Moropoulou · Monika Hagedorn-Saupe

Antonella Fresa · Gunnar Liestøl
Vlatka Rajcic · Pierre Grussenmeyer (Eds.)
Digital Heritage
LNCS 10058
Progress in Cultural Heritage:

Documentation, Preservation,
and Protection
6th International Conference, EuroMed 2016
Nicosia, Cyprus, October 31 – November 5, 2016
Proceedings, Part I
123
Lecture Notes in Computer Science 10058
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen
Editorial Board
David Hutchison
Lancaster University, Lancaster, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Friedemann Mattern
ETH Zurich, Zurich, Switzerland
John C. Mitchell
Stanford University, Stanford, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
C. Pandu Rangan
Indian Institute of Technology, Madras, India
Bernhard Steffen
TU Dortmund University, Dortmund, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany
More information about this series at http://www.springer.com/series/7409
Marinos Ioannides Eleanor Fink
•
Antonia Moropoulou Monika Hagedorn-Saupe

•
Antonella Fresa Gunnar Liestøl

•
Vlatka Rajcic Pierre Grussenmeyer (Eds.)

•
Digital Heritage
Progress in Cultural Heritage:
Documentation, Preservation,
and Protection
6th International Conference, EuroMed 2016
Nicosia, Cyprus, October 31 – November 5, 2016
Proceedings, Part I
123
Editors
Marinos Ioannides Antonella Fresa
Cyprus University of Technology Promoter s.r.l.
Limassol Peccioli
Cyprus Italy
Eleanor Fink Gunnar Liestøl
Arlington, VI University of Oslo
USA Oslo
Norway
Antonia Moropoulou
National Technical University of Athens Vlatka Rajcic
Athens University of Zagreb
Greece Zagreb
Croatia
Monika Hagedorn-Saupe
Institut für Museumsforschung Pierre Grussenmeyer
Berlin INSA
Germany Strasbourg
France
Cover illustration: The illustration appearing on the cover of this book is used with permission.
Copyright holder: Byzantine Museum and Art Galleries, Archbishop Makarios III Foundation,
Arch. Kyprianos sqr., P.O. Box 21269, 1505 Nicosia, Cyprus. http://www.makariosfoundation.org.cy
ISSN 0302-9743 ISSN 1611-3349 (electronic)

Lecture Notes in Computer Science
ISBN 978-3-319-48495-2 ISBN 978-3-319-48496-9 (eBook)
DOI 10.1007/978-3-319-48496-9
Library of Congress Control Number: 2016956494
LNCS Sublibrary: SL3 – Information Systems and Applications, incl. Internet/Web, and HCI
© Springer International Publishing AG 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are
believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors
give a warranty, express or implied, with respect to the material contained herein or for any errors or
omissions that may have been made.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
Welcome to the proceedings of EuromedMed 2016, the biennial scientific event which
this year was held in the capital city of Cyprus, the island that has always been a bridge
to three continents in the world going back to the origins of civilization. It is a place
where the fingerprints of several ancient cultures and civilizations on earth can be
found, with a wealth of historical sites recognized and protected by UNESCO.
Several organizations and current EU projects (such as the Marie Sklodowska-Curie
Fellowship project on Digital Heritage Marie Sklodowska-Curie FP7-PEOPLE
ITN-DCH, the Marie Sklodowska-Curie FP7-IAPP 4D-CH-WORLD, the FP7-CIP
ICT-PSP EuropeanaSpace, the H2020 Reflective 7 - INCEPTION, the H2020 CSA
Virtual Museums ViMM, the Research Infrastructure DARIAH-EU ERIC and
DARIAH-CY) as well as the Innovation in Intelligent Management of Heritage
Buildings (i2MHB) decided to join EuroMed2016 and continue cooperating together in
order to create an optimal environment for the discussion and explanation of new
technologies, the exchange of modern innovative ideas, and in general to allow the
transfer of knowledge between a large number of professionals and academics during
one common event.
The main goal of the event is to illustrate the programs underway, whether orga-
nized by public bodies (e.g., UNESCO, European Union, National States, etc.) or by
private foundations (e.g., Getty Foundation, World Heritage Foundation, etc.) in order
to promote a common approach to the tasks of recording, documenting, protecting, and
managing world cultural heritage. The 6th European-Mediterranean Conference
(EuroMed 2016) was definitely a forum for sharing views and experiences, discussing
proposals for the optimum approach as well as the best practice and the ideal technical
tools to preserve, document, manage, present/visualize and disseminate the rich and
diverse cultural heritage of mankind.
This conference was held during the mid-term of the new Framework Programme,
Horizon 2020, which is the largest in the world in terms of financial support on
research, innovation, technological development, and demonstration activities. The
awareness of the value and importance of heritage assets has been reflected in the
financing of projects since the first Framework Programme for Research & Techno-
logical Development (FP1, 1984–87) and continues into current HORIZON 2020 that
follows FP7 (2007–13). In the past 30 years, a large community of researchers, experts,
and specialists have had the chance to learn and develop the transferable knowledge
and skills needed to inform stakeholders, scholars, and students. Europe has become a
leader in heritage documentation, preservation, and protection science, with COST
Actions adding value to projects financed within the FP and EUREKA programme and
transferring knowledge to practice and supporting the development of SMEs.
The EuroMed 2016 agenda focused on enhancing and strengthening of international
and regional cooperation and promoting awareness and tools for future innovative
research, development, and applications to protect, preserve, and document the
VI Preface
European and world cultural heritage. Our ambition was to host an exceptional con-
ference by mobilizing also policy makers from different EU countries, institutions
(European Commission, European Parliament, Council of Europe, UNESCO, Inter-
national Committee for Monuments and Sites ICOMOS, the International Committee
for Documentation of Cultural Heritage CIPA, the International Society for Pho-
togrammetry and Remote Sensing ISPRS, the International Centre for the study of the
Preservation and Restoration of Cultural Property ICCROM, and the International
Committee for Museums ICOM), professionals, as well as participants from all over
the world and from different scientific areas of cultural heritage.
Protecting, preserving, and presenting our cultural heritage are actions that are
frequently interpreted as change management and/or changing the behavior of society.
Joint European and international research produce the scientific background and sup-
port for such a change. We are living in a period characterized by rapid and remarkable
changes in the environment, in society, and in technology. Natural changes, war
conflicts, and man-made changes, including climate, as well as technological and
societal changes, form an ever-moving and colorful stage and a challenge for our
society. Close cooperation between professionals, policy makers, and authorities
internationally is necessary for research, development, and technologica advancements
in the field of cultural heritage.
Scientific projects in the area of cultural heritage have received national, European
Union, or UNESCO funding for more than 30 years. Through financial support and
cooperation, major results have been achieved and published in peer-reviewed journals
and conference proceedings with the support of professionals from many countries.
The European Conferences on Cultural Heritage research and development and in
particular the biennial EuroMed conference have become regular milestones on the
never-ending journey in the search for new knowledge of our common history and its
protection and preservation for the generations to come. EuroMed also provides a
unique opportunity to present and review results as well as to draw new inspiration.
To reach this ambitious goal, the topics covered include experiences in the use of
innovative technologies and methods and how to take best advantage to integrate the
results obtained to build up new tools and/or experiences as well as to improve
methodologies for documenting, managing, preserving, and communicating cultural
heritage.
In these proceedings we present 105 papers, selected from 504 submissions, which
focus on interdisciplinary and multidisciplinary research concerning cutting-edge cul-
tural heritage informatics, physics, chemistry, and engineering and the use of tech-
nology for the representation, documentation, archiving, protection, preservation, and
communication of cultural heritage knowledge.
Our Keynote speakers, Prof. Dr. Antonia Moropoulou (NTUA and Technical
Chamber of Greece), Prof. Dr. Dieter Fellner (Director of FhD/IGD and TU Darmstadt,
Germany), Prof. Dr. Wolfgang Kippes (University for Applied Arts Vienna and Donau
University Krems, Austria), Prof. Dr. Sarah Whatley (Director of Centre for Dance
Research, UK), Prof. Dr. Mustafa Erdik (Bogazici University of Instabul, Turkey), Mr.
Jean-Pierre Massué (Senate Member of the European Academy of Sciences and Arts/
COPRNM, France), Mr. Axel Ermert (Institute for Museum Research SMB/PK of
Berlin, Germany), Mrs. Rosella Caffo (Director of the Central Institute for the Union
Preface VII
Catalogue of the Italian Libraries (ICCU), Italy), Mr. Vasco Fassina (President of the
European Standardization Commission CEN/TC 346: Conservation of Cultural Her-
itage, Italy), Mrs. Maria P. Kouroupas (Director Cultural Heritage Center, US
Department of State), Mrs. France Desmarais (ICOM), Dr. Thomas R. Klein (Counsel,
Andrews Kurth LLP), Françoise Bortolotti (Criminal Intelligence Officer, Works of Art
Unit, Interpol) and Prof. Dr. Markus Hilgert (Director, Vorderasiatisches Museum im
Pergamonmuseum Staatliche Museen zu Berlin - Preußischer Kulturbesitz and Project
Leader, ILLICID) are not only experts in their fields, but also visionaries for the future
of cultural heritage protection and preservation. They promote the e-documentation and
protection of the past in such a way for its preservation for the generations to come.
We extend our thanks to all authors, speakers, and those persons whose labor,
financial support, and encouragement made the EuroMed 2016 event possible. The
International Program Committee—whose members represent a cross-section of
archaeology, physics, chemistry, civil engineering, computer science, graphics and
design, library, archive and information science, architecture, surveying, history and
museology—worked tenaciously and finished their work on time. The staff of the IT
department at the Cyprus University of Technology helped with their local ICT and
audio visual support, especially Mr. Filippos Filippou, Mr. Costas Christodoulou, and
Mr. Stephanos Mallouris. We would also like to express our gratitude to all the
organizations supporting this event and our co-organizers, the European Commission
scientific and policy officers of the H2020 Marie Skłodowska-Curie Programme, the
director general of Europeana, Mrs. Jill Cousins, the Getty Conservation Institute and
World Monuments Fund, the Cyprus University of Technology, the Ministry of
Energy, Commerce, Industry, and Tourism. Especially the permanent secretary and
Digital Champion Dr. Stelios Himonas and Mr. Nikos Argyris, the Ministry of Edu-
cation and Culture and particularly Minister Dr. Costas Kadis, the director of Cultural
Services Mr. Pavlos Paraskevas, the Department of Antiquities in Cyprus, all the
members of the Cypriot National Committee for E-Documentation and E-Preservation
in Cultural Heritage, and finally our corporate sponsors, CableNet Ltd., the Cyprus
Tourism Organization, the Cyprus Postal Services, the Cyprus Handicraft Center, and
Dr. Kyriacos Themistocleous from the Cyprus Remote Sensing Society, who provided
services and gifts in kind that made the conference possible.
We express our thanks and appreciation to Dr. Nikos Grammalides from CERTH in
Greece and Dr. Sander Münster, the Dresden University of Technology, Germany, as
well as the board of the ICOMOS Cyprus Section for their enthusiasm, commitment,
and support for the success of this event. Most of all we would like to thank the
organizations UNESCO, European Commission, CIPA, ISPRS, and ICOMOS Europa
Nostra that entrusted us with the task of organizing and undertaking this unique event.
September 2016 Marinos Ioannides

Eleanor Fink
Antonia Moropoulou
Antonella Fresa
Gunnar Liestøl
Vlatka Rajcic
Pierre Grussenmeyer
Acknowledgments and Disclaimer
The EuroMed 2016 conference was partly supported by the Republic of Cyprus, the
Cyprus University of Technology, the Cyprus Tourism Organization, the CableNet Ltd.,
by CIPA (http://cipa.icomos.org/), ISPRS, ICOMOS, Europa Nostra the EU projects
FP7 PEOPLE ITN2013 ITN-DCH and IAPP2012 4D-CH-WORLD, the DARIAH-EU
ERIC and DARIAH-CY, the FP7-ICT-2011 i-Treasures, the CIP ICT-PSP Euro-
peanaSpace and H2020 INCEPTION and H2020-ViMM projects.
However, the content of this publication reflects only the authors’ views; the European
Commission, the Republic of Cyprus, CIPA, ISPRS, ICOMOS, Europa Nostra, Cyprus
University of Technology and the EU projects FP7 PEOPLE ITN2013 ITN-DCH and
IAPP2012 4D-CH-WORLD, the DARIAH-EU ERIC and DARIAH-CY, the
FP7-ICT-2011 i-Treasures, the CIP ICT-PSP EuropeanaSpace and H2020-INCEPTION
and H2020-ViMM are not liable for any use that may be made of the information con-
tained herein.
Organization
Conference Chairs
Marinos Ioannides
Eleanor Fink
Antonia Moropoulou
Antonella Fresa
Gunnar Liestøl
Vlatka Rajcic
Pierre Grussenmeyer
Paper Review Chair
Pavlos Chatzigrigoriou
Local Organizing Committee
Agapiou, Athos Nobilakis, Elias

Chatzigrigoriou, Pavlos Papageorgiou, Eirini
Eliades, Ioannis Skriapas, Konstantinos
Gkanetsos, Theodoros Yianni, Stephanie
Leventis, Georgios Stylianou, Georgios
Marina, Christodoulou Themistocleous, Kyriacos
Nikolakopoulou, Vasiliki Athanasiou, Vasilios
International Scientific Committee

Agapiou, Athos Cyprus Caliandro, Rocco Italy
Albertson, Lynda Italy Callet, Patrick Italy
Angeletaki, Alexandra Norway Chatzigrigoriou, Pavlos Greece
Balet, Olivier France Colla, Camilla Italy
Barcelo, Juan Spain Corsi, Cristina Italy
Bebis, George USA Cuca, Branca Serbia
Belgiorno, Maria-Rosaria Italy Dallas, Costis Canada
Bellini, Francesco Italy De Jong, Annemieke Netherlands
Berni, Marco Italy De Kramer, Marleen Austria
Bockholt, Ulrich Germany De Leeuw, Sonja Netherlands
Bryan, Paul UK De Masi, Alessandro Italy
X Organization
De Niet, Marco Netherlands Liestøl, Gunnar Norway

De Polo Saibanti, Andrea Italy Lobovikov Katz, Anna Israel
Degraeve, Ann Belgium Lonnqvist, Minna Lichtenstein
Dobreva, Milena Malta Lopez-Menchero Bendicho, Italy
Domajnko, Matevz Slovenia Victor
Doneus, Michael Austria Madija, Lidija Serbia
Doulamis, Anastasios Greece Maietti, Federica Italy
Doulamis, Nikolaos Greece Makantasis, Konstantinos Greece
Drap, Pierre France Marcella, Stefano Italy
Eppich, Rand USA Martins, Joao Portugal
Farrag, Maged Egypt Masini, Nicola Italy
Filin, Sagi Israel Mate, Toth Hungary
Fouseki, Kalliopi UK Michail, Harris Cyprus
Fresa, Antonella Italy Moropoulou, Antonia Greece
Frick, Jürgen Germany Munster, Sander Germany
Gebhardt, Andreas Germany Nanetti, Andrea Singapore
Giannoulopoulos, Spain Nikolakopoulou, Vasiliki Greece
Giannoulis Georgios Nurminen, Antti Finland
Giuliano, Angele Italy Nys, Karin Belgium
Graf, Holger Germany Ouimet, Christian Canada
Grammalidis, Nikos Greece Papageorgiou, Dimitris Greece
Grosset, Marie France Papageorgiou, Eirini Greece
Grussenmeyer, Pierre France Papagiannakis, George Greece
Gutierrez, Mariano Flores Spain Petrelli, Daniela UK
Gutiérrez Meana, Javier Spain Pietro, Liuzzo Germany
Hagedorn-Saupe, Monika Germany Potsiou, Chryssy Greece
Hanke, Klaus Austria Protopapadakis, Eftychios Greece
Ibáñez, Francisco Spain Radoslav, Pavlov Bulgary
Ioannidis, Charalambos Greece Rajcic, Vlatka Hungary
Jabi, Wassim UK Rodriguez-Echavarria, UK
Kersten, Thomas Germany Karina
Klein, Michael Austria Roko, Zarnic Slovenia
Klein, Reinhard Germany Ronchi, Alfredo Italy
Kolias, Stefanos Greece Saleh, Fathi Egypt
Koukios, Emmanuel Greece Sánchez Andreu, Joan Spain
Koutsabasis, Panayiotis Greece Santana, Mario Canada
Kuroczyński, Piotr Germany Santos, Pedro Germany
Kyriakaki, Georgia Greece Schindler, Mathias Germany
Landes, Tania France Sempere, Isabel Martínez Spain
Lange, Guus Belgium Shang, Jin China
Laquidara, Giuseppe Italy Snyders, Marius Netherlands
Leissner, Johanna Germany Stork, Andre Germany
León, Alfredo Grande Spain Tapinaki, Sevasti Greece
Lerma, José Luis Spain Themistocleous, Kyriacos Cyprus
Leventis, Georgios Greece Thwaites, Harold MY
Liarokapis, Fotis Greece Tsai, Fuan Taiwan
Organization XI
Tsapatsoulis, Nicolas Cyprus Vermeulen, Frank Belgium

Tsiafaki, Despoina Greece Vosinakis, Spyros Greece
Tsoupikova, Daria USA Walczak, Krzysztof Poland
Uueni, Andres Estonia Wallace, Manolis Greece
Vander Vorst, Daniel Spain Wehr, Aloysius Germany
Vassilara, Archontoula Greece Wilk, Christian Germany
Vavalis, Manolis Greece Yen, Alex Taiwan
The Icons of the Chapel of Saint Jacob
The icon shown on the cover of LNCS 10058 (Part I) depicts the scene of the
Enthroned Virgin Mary with Child together with Saint John the Evangelist, while the
icon shown on the cover of LNCS 10059 (Part II) illustrates Jesus Christ on a throne
together with Saint John the Baptist. The icons are dated back to 1620 A.D. and were
painted by the artist Meletios from Crete. These icons were stolen from the iconostasis
of the chapel of Saint Jacob in Trikomo (Famagusta district) after the Turkish invasion
of 1974. Saint Jacob’s chapel had no frescoes but it was decorated with colorful plates
of traditional folk art.
The icon illustrating Jesus Christ and Saint John the Baptist is 110 128 cm in size
and close to the feet of the latter there is the inscription “XEIP MEKETIO! TO!
KPITOR AXK(= 1620) X(qirso) !. M(ηmo1) atcoύrsot),” which includes the name
of the artist as well as the date. The icon of Mary, Mother of Jesus, together with Saint
John the Evangelist is 114 134 cm in size. Both of them were in the possession of the
Russian–Jewish art dealer Alexander Kocinski, until their confiscation by the Swiss
Police in Zurich in 2007. The only documentation available to recover these stolen
icons from abroad was a paper published in the Proceedings of the International
Cretan Conference in 1976 by the former director of the Department of Antiquities of
Cyprus, Mr. Athanasios Papageorgiou.
The icons were tracked down in 2007 in Christie’s Auction House in London, from
where they were withdrawn after actions by Kykkos Monastery. Following information
by the bishop of Kykkos Monastery, representatives of the monastery traveled to Zurich
to meet the owner of the icons; however, it was not possible to persuade him to return the
icons to the lawful owners and therefore the authorities of Cyprus were informed.
A written complaint by the Byzantinologist of Kykkos Monastery, Dr. C. Chotzakoglou,
to the Cypriot Police and to Interpol in Cyprus initiated the repatriation procedure of the
icons, eventually leading to their confiscation by the Swiss Interpol.
By means of a new testimony from Dr. C. Chotzakoglou, in addition to a full
documentation of the Cypriot origin of the icons and their looting after the Turkish
invasion in northern Cyprus, the Supreme Court of Famagusta, based in Larnaca, took
legal measures against the owner of the icons, who was convicted. The verdict of the
Cypriot Supreme Court was subsequently used in the Swiss Court, leading to the
signing of a compromise settlement between the Church of Cyprus and Kocinski for
the return of the icons to Cyprus.
Dr. Ioannis A. Eliades,

Director,
Byzantine Museum and Art Galleries,
Archbishop Makarios III Foundation,
Arch. Kyprianos sqr., P.O. Box 21269,
CY-1505 Nicosia, CYPRUS
Contents – Part I
Full Paper: 3D Reconstruction and 3D Modelling
Implementation and Usage Scenarios of a Participatory Platform

for Cultural Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Zois Koukopoulos and Dimitrios Koukopoulos
Benchmarking Close-range Structure from Motion 3D Reconstruction

Software Under Varying Capturing Conditions . . . . . . . . . . . . . . . . . . . . . . 15
Ivan Nikolov and Claus Madsen
Proportional Systems in the Design of the Cathedral of St. George

of the Greeks, Cyprus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Douglas Cawthorne and Romylos Irodotou
The Reconstruction – Argumentation Method: Proposal for a Minimum

Standard of Documentation in the Context of Virtual Reconstructions . . . . . . 39
Mieke Pfarr-Harfst and Marc Grellert
Multi-scale 3D Modelling of Damaged Cultural Sites: Use Cases

and Image-Based Workflows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Styliani Verykokou, Anastasios Doulamis, George Athanasiou,
Charalabos Ioannidis, and Angelos Amditis
Low Cost Technique for Accurate Geometric Documentation of Complex

Monuments by Non-experts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Charalabos Ioannidis, Sofia Soile, and Styliani Verykokou
Full Paper: Heritage Building Information Models (HBIM)
Implementation of Scan-to-BIM and FEM for the Documentation

and Analysis of Heritage Timber Roof Structures . . . . . . . . . . . . . . . . . . . . 79
Maarten Bassier, George Hadjidemetriou, Maarten Vergauwen,
Nathalie Van Roy, and Els Verstrynge
Implementation Analysis and Design for Energy Efficient Intervention

on Heritage Buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Elena Gigliarelli, Filippo Calcerano, and Luciano Cessari
Historic BIM in the Cloud . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

Luigi Barazzetti, Fabrizio Banfi, and Raffaella Brumana
XVI Contents – Part I
Building Information Modelling – A Novel Parametric Modeling Approach

Based on 3D Surveys of Historic Architecture . . . . . . . . . . . . . . . . . . . . . . 116
Fabrizio Banfi
Full Paper: Innovative Methods on Risk Assesment, Monitoring

and Protection of Cultural Heritage
Applicability of 3DVE Indicator for Cultural Heritage Landscape

Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Jaeyong Lee and Youngmo Kim
Taking the Next Step in Digital Documentation of Historic Cities:

How HERMeS Evolved in an Open Data Digital Library of Historic
Buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Pavlos Chatzigrigoriou
Risk Analysis and Vulnerability Assessment of Archeological Areas

for the Preventive Conservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Giuliana Quattrone
Full Paper: Intangible Cultural Heritage Documentation
Parameterizing the Geometry and Visualizing the Lighting Method

of Byzantine Church Domes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Wassim Jabi and Iakovos Potamianos
Digital and Handcrafting Processes Applied to Sound-Studies

of Archaeological Bone Flutes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Etienne Safa, Jean-Baptiste Barreau, Ronan Gaugne,
Wandrille Duchemin, Jean-Daniel Talma, Bruno Arnaldi,
Georges Dumont, and Valérie Gouranton
Full Paper: Digital Applications for Materials’ Preservation

and Conservation in Cultural Heritage
Evolution of Building Materials and Philosophy in Construction: A Process

of Digitalization and Visualization of the Accumulated Knowledge . . . . . . . . 199
Ioanna Papayianni and Vasiliki Pachta
A Study of 3D Digital Simulation Analysis of Fire Charring Degree

of Wood Construction of Chinese Traditional Architecture . . . . . . . . . . . . . . 209
Tsung Chiang Wu
Contents – Part I XVII
Full Paper: Non-destructive Techniques in Cultural Heritage Conservation
Recovering Historical Film Footage by Processing

Microtomographic Images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Chang Liu, Paul L. Rosin, Yu-Kun Lai, Graham R. Davis, David Mills,
and Charles Norton
Multi-spectral Imaging System (IWN) for the Digitization and Investigation

of Cultural Heritage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Ibrahim El-Rifai, Hend Mahgoub, and Ari Ide-Ektessabi
Diagnostic Activities for the Planned and Preventive Conservation

of Mosaic Pavements: The Case Study of the Triclinium
of the Villa Romana del Casale (Sicily) . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Antonella Versaci, Alessio Cardaci, and Luca R. Fauzia
Full Paper: Visualisation, VR and AR Methods and Applications
A Mobile, AR Inside-Out Positional Tracking Algorithm, (MARIOPOT),

Suitable for Modern, Affordable Cardboard-Style VR HMDs . . . . . . . . . . . . 257
Paul Zikas, Vasileios Bachlitzanakis, Margarita Papaefthymiou,
and George Papagiannakis
Virtual and Augmented Reality Tools to Improve the Exploitation

of Underwater Archaeological Sites by Diver and Non-diver Tourists . . . . . . 269
Fabio Bruno, Antonio Lagudi, Loris Barbieri, Maurizio Muzzupappa,
Gerardo Ritacco, Alessandro Cozza, Marco Cozza, Raffaele Peluso,
Marco Lupia, and Gianni Cario
Interacting with Simulated Archaeological Assets . . . . . . . . . . . . . . . . . . . . 281

Arian Goren, Kay Kohlmeyer, Thomas Bremer, Susanne Brandhorst,
Arie Kai-Browne, Felix Balda, David Strippgen, and Sebastian Plesch
Virtual Reconstruction 3.0: New Approach of Web-based Visualisation

and Documentation of Lost Cultural Heritage . . . . . . . . . . . . . . . . . . . . . . . 292
Daniel Dworak and Piotr Kuroczyński
Full Paper: The New Era of Museums and Exhibitions: Digital Engagement
and Dissemination
CultureCam: An Interactive Search Tool for Small Image Galleries . . . . . . . . 309

Sergiu Gordea, Michela Vignoli, and Sanna Marttila
Learning Algorithms for Digital Reconstruction of Van Gogh’s Drawings . . . 322

Yuan Zeng, Jiexiong Tang, Jan C.A. van der Lubbe, and Marco Loog
XVIII Contents – Part I
Full Paper: Digital Cultural Heritage in Education, Learning and Training
“Human” Technology in the Digital Era: Freehand Images and Analysis

of Cultural Heritage – The Know-How and Its Applications . . . . . . . . . . . . . 337
Anna Lobovikov-Katz
Adult and Children User Experience with Leap Motion in Digital Heritage:
The Cycladic Sculpture Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Panayiotis Koutsabasis and Spyros Vosinakis
Researching Knowledge Concerns in Virtual Historical Architecture . . . . . . . 362

S. Münster, C. Kröber, H. Weller, and N. Prechtel
Project Paper: Data Acquisition, Process and Management

in Cultural Heritage
Surveying Illusory Architectures Painted on Vaulted Surfaces. . . . . . . . . . . . 377

Matteo Flavio Mancini and Marta Salvatore
The Application of Photogrammetry on Digitization and Promotion

for Monuments and Temples in Taiwan - Taking Chua Family
Ancestral Temple as an Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Wun-Bin Yang, Tsung-Juang Wang, and Ya-Ning Yen
3D Acquisition, Processing and Visualization of Archaeological Artifacts:

The Samarra Collection of the Museum of Islamic Art in Berlin . . . . . . . . . . 397
Arie Kai-Browne, Kay Kohlmeyer, Julia Gonnella, Thomas Bremer,
Susanne Brandhorst, Felix Balda, Sebastian Plesch,
and Dennis Lehmann
PHOTOCONSORTIUM: Digitizing Europe’s Photographic Heritage . . . . . . . 409

Frederik Truyen and Antonella Fresa
Acquisition and Processing Experiences of Close Range UAV Images

for the 3D Modeling of Heritage Buildings . . . . . . . . . . . . . . . . . . . . . . . . 420
Arnadi Murtiyoso, Pierre Grussenmeyer, Mathieu Koehl,
and Tristan Freville
Internal 3D Printing of Intricate Structures . . . . . . . . . . . . . . . . . . . . . . . . . 432

Théophane Nicolas, Ronan Gaugne, Cédric Tavernier,
Valérie Gouranton, and Bruno Arnaldi
Towards Monuments’ Holistic Digital Documentation: the Saint Neophytos

Enkleistriotis Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
Marinos Ioannides, Charalabos Ioannidis,
Archimadrite Neophytos Enkleistriotis, David Castrillo,
Pavlos Chatzigrigoriou, Eirini Papageorgiou, Georgios Leventis,
Vasiliki Nikolakopoulou, Vasilis Athanasiou, Fotis Bourexis, Sofia Soile,
Styliani Verykokou, Maria Costi de Castrillo, and Christian Sovis
Contents – Part I XIX
Project Paper: Data, Metadata, Semantics and Ontologies in Cultural Heritage
First Experiences of Applying a Model Classification for Digital 3D

Reconstruction in the Context of Humanities Research. . . . . . . . . . . . . . . . . 477
Sander Münster, Cindy Kröber, Wolfgang Hegel, Mieke Pfarr-Harfst,
Nikolas Prechtel, Rainer Uhlemann, and Frank Henze
Digital Preservation of Cultural Heritage: Balinese Kulkul Artefact

and Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
Cokorda Pramartha and Joseph G. Davis
Interconnecting Objects, Visitors, Sites and (Hi)Stories Across Cultural

and Historical Concepts: The CrossCult Project . . . . . . . . . . . . . . . . . . . . . 501
Costas Vassilakis, Angeliki Antoniou, George Lepouras,
Manolis Wallace, Ioanna Lykourentzou, and Yannick Naudet
Project Paper: 3D Reconstruction and 3D Modelling
Debate and Considerations on Using Videos for Cultural Heritage

from Social Media for 3D Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
Kyriacos Themistocleous
Investigations of Low-Cost Systems for 3D Reconstruction

of Small Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
Thomas P. Kersten, Daniel Omelanowsky, and Maren Lindstaedt
Digital Documentation: Villa Borghese . . . . . . . . . . . . . . . . . . . . . . . . . . . 533

Martina Attenni, Cristiana Bartolomei, and Alfonso Ippolito
Digital 3D Reconstructed Models – Structuring Visualisation

Project Workflows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544
Mieke Pfarr-Harfst and Stefanie Wefers
Reconstruction of Wooden “Polish Manor” . . . . . . . . . . . . . . . . . . . . . . . . 556

Anna Rozanska and Wojciech Koryciński
Project Paper: Heritage Building Information Models (HBIM)
Digital Tools for Heritage Preservation and Enhancement: The Integration

of Processes and Technologies on 20th Century Buildings in Brazil
and India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
Luca Rossato
From Integrated Survey to the Parametric Modeling of Degradations.

A Feasible Workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Massimiliano Lo Turco, Federico Caputo, and Gabriele Fusaro
XX Contents – Part I
INCEPTION Standard for Heritage BIM Models. . . . . . . . . . . . . . . . . . . . . 590

Peter Bonsma, Iveta Bonsma, Anna Elisabetta Ziri, Silvia Parenti,
Pedro Martín Lerones, José Luis Hernández, Federica Maietti,
Marco Medici, Beatrice Turillazzi, and Ernesto Iadanza
From SfM to Semantic-Aware BIM Objects of Architectural Elements . . . . . 600

Massimiliano Lo Turco and Cettina Santagati
Project Paper: Novel Approaches to Landscapes in Cultural Heritage
Observing Landscape Changes Around the Nicosia Old Town Center

Using Multi-temporal Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
Branka Cuca, Athos Agapiou, and Diofantos G. Hadjimitsis
Towards the Sustainable Development of Cultural Landscapes

Through Two Case Studies on Different Scale . . . . . . . . . . . . . . . . . . . . . . 625
Eirini Papageorgiou
The Fortification System of the City of Urbino: The Case Study of Valbona
Gate from 3D Surveys to GIS Applications for Dynamic Maps. . . . . . . . . . . 645
Sara Bertozzi, Laura Baratin, and Elvio Moretti
Project Paper: Innovative Methods on Risk Assessment, Monitoring

and Protection of Cultural Heritage
Disaster-Risk Indicators and Their Evaluation, Communication

for Traditional Settlement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Alex Ya-Ning Yen and Chin-Fang Cheng
Pervasive Wireless Sensor Networks for the Monitoring of Large

Monumental Structures: The Case of the Ancient City Walls of Siena . . . . . . 669
Alessandro Pozzebon, Tommaso Addabbo, Ada Fort, Marco Mugnaini,
Enza Panzardi, and Valerio Vignoli
Project Paper: Digital Applications for Materials’ Preservation

and Conservation in Cultural Heritage
The SACRE Project: A Diagnosis Tool of Built Heritage. . . . . . . . . . . . . . . 681

Sarah Janvier-Badosa, Kévin Beck, Xavier Brunetaud,
and Muzahim Al-Mukhtar
Pigments Identification Using Raman Spectroscopy of the 16th Century

Printed Book “Osorio”. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
Igor Lukačević, Theodore Ganetsos, and Thomas Katsaros
Contents – Part I XXI
Design and Application of a Data System for the Comparative Study

of Historic Mortars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
Vasiliki Pachta and Ioanna Papayianni
GIS Applications for a New Approach to the Analysis of Panel Paintings . . . 711
Laura Baratin, Sara Bertozzi, Elvio Moretti, and Roberto Saccuman
Project Paper: Visualisation, VR and AR Methods and Applications
‘Translation’ and Fruition of an Ancient Book Through Virtual Reality

in the Case of Lost Cultural Heritage: Case Study: “Inscriptiones”
by Emmanuel Thesaurus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727
Sanaz Davardoust, Anna Osello, and Rosa Tamborrino
An Interdisciplinary Study on the Ancient Egyptian Wines:

The Egywine Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
Maria Rosa Guasch-Jané
ArchaeoInside: Multimodal Visualization of Augmented Reality

and Interaction with Archaeological Artifacts . . . . . . . . . . . . . . . . . . . . . . . 749
Kadar Manuella and Domsa Ovidiu
DICE: Digital Immersive Cultural Environment . . . . . . . . . . . . . . . . . . . . . 758

Stelios C.A. Thomopoulos, Adam Doulgerakis, Maria Bessa,
Konstantinos Dimitros, Giorgos Farazis, Eftichia Georgiou,
Tassos Kanellos, Christina Karafylli, Maria Karafylli,
Dimitris M. Kyriazanos, Vassilios I. Kountouriotis,
Vassilis Lampropoulos, Christos Margonis, Christos Maroglou,
Dionisis Motos, Alexandra Papagianni, Manolis Paterakis,
Katerina Skroumpelou, Giorgos Konstandinos Thanos,
Ino-Eleni Theodorou, Christina Phobe Thomopoulos,
Panagiotis Tsimpiridis, Dimitris Zacharakis, and Andreas Zalonis
Project Paper: The New Era of Museums and Exhibitions: Digital

Engagement and Dissemination
Imaging Novecento. A Mobile App for Automatic Recognition of Artworks

and Transfer of Artistic Styles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781
Federico Becattini, Andrea Ferracani, Lea Landucci, Daniele Pezzatini,
Tiberio Uricchio, and Alberto Del Bimbo
Towards the Design of a User-Friendly and Trustworthy Mobile System

for Museums . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792
Kostas Koukoulis and Dimitrios Koukopoulos
XXII Contents – Part I
Project Paper: Serious Games for Cultural Heritage
Project iMARECULTURE: Advanced VR, iMmersive Serious Games

and Augmented REality as Tools to Raise Awareness and Access
to European Underwater CULTURal heritagE. . . . . . . . . . . . . . . . . . . . . . . 805
D. Skarlatos, P. Agrafiotis, T. Balogh, F. Bruno, F. Castro,
B. Davidde Petriaggi, S. Demesticha, A. Doulamis, P. Drap,
A. Georgopoulos, F. Kikillos, P. Kyriakidis, F. Liarokapis, C. Poullis,
and S. Rizvic
Immersivity and Playability Evaluation of a Game Experience

in Cultural Heritage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
Roberto Andreoli, Angela Corolla, Armando Faggiano,
Delfina Malandrino, Donato Pirozzi, Mirta Ranaldi,
Gianluca Santangelo, and Vittorio Scarano
Tirolcraft: The Quest of Children to Playing the Role of Planners

at a Heritage Protected Town . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825
Bruno Amaral de Andrade, Ítalo Sousa de Sena,
and Ana Clara Mourão Moura
Pervasive Game Utilizing WiFi Fingerprinting-based Localization . . . . . . . . . 836

Filip Maly, Pavel Kriz, and Michael Adamec
Project Paper: Digital Cultural Heritage in Education, Learning and Training
ErfgoedApp: An Educational Experiment with Augmented Reality,

Cityscapes and Campusscapes in Brussels . . . . . . . . . . . . . . . . . . . . . . . . . 849
Marc Jacobs, Morien Schroyen, and Joke Vanderschoot
Contextualizing 3D Cultural Heritage . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859

James Lloyd
Crowdsourcing Cultural Heritage: From 3D Modeling to the Engagement

of Young Generations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869
Laura Inzerillo and Cettina Santagati
Training Schools for Conservation of Cultural Heritage: Between Expertise,

Management and Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
Anna Lobovikov-Katz, Gumersindo Bueno Benito,
Vanesa Marcos Sánchez, Joao Martins, and Dalik Sojref
Educational Creative Use and Reuse of Digital Cultural Heritage Data

for Cypriot UNESCO Monuments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
Marinos Ioannides, Pavlos Chatzigrigoriou, Vasilis Bokolas,
Vasiliki Nikolakopoulou, Vasilis Athanasiou, Eirini Papageorgiou,
Georgios Leventis, and Christian Sovis
Contents – Part I XXIII
Time-Travelling with Mobile Augmented Reality: A Case Study

on the Piazza dei Miracoli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902
Mihai Duguleana, Raffaello Brodi, Florin Girbacia, Cristian Postelnicu,
Octavian Machidon, and Marcello Carrozzino
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913

Contents – Part II
Data Acquisition, Process and Management in Cultural Heritage
Development of Photogrammetric Documentation of the Borough

at Biskupin Based on Archival Photographs - First Results. . . . . . . . . . . . . . 3
D. Zawieska and J. Markiewicz
Capturing Our Cultural Intangible Textile Heritage, MoCap

and Craft Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Eva Andersson Strand, Stefan Lindgren, and Carolina Larsson
CH Digital Documentation and 3D Survey to Foster the European

Integration Process: The Case Study of Geguti Palace in Kutaisi, Georgia . . . 16
Marco Medici, Federico Ferrari, Nana Kuprashvili, Tamar Meliva,
and Nino Bugadze
Data, Metadata, Semantics and Ontologies in Cultural Heritage
Applying Deep Learning Techniques to Cultural Heritage Images

Within the INCEPTION Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Jose Llamas, Pedro M. Lerones, Eduardo Zalama,
and Jaime Gómez-García-Bermejo
Reaching the World Through Free Licenses and Wikimedia’s

Crowdsourced Platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
John Andersson
3D Reconstruction and 3D Modelling
Definition of a Workflow for Web Browsing of 3D Models in Archaeology. . . 41

A. Scianna, M. La Guardia, and M.L. Scaduto
The VALMOD Project: Historical and Realistic 3D Models for the Touristic
Development of the Château de Chambord. . . . . . . . . . . . . . . . . . . . . . . . . 53
Xavier Brunetaud, Romain Janvier, Sarah Janvier-Badosa, Kévin Beck,
and Muzahim Al-Mukhtar
Novel Approaches to Landscapes in Cultural Heritage
Geological Heritage and Conservation: A Case Study of the Visual

Axis Through Digital Terrain Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Pedro Casagrande, Nicole Rocha, Ítalo Sena, Bráulio Fonseca,
and Ana Clara Moura
XXVI Contents – Part II
A GIS Database of Montenegrin Katuns (Kuči Mountain and Durmitor) . . . . 72

Olga Pelcer–Vujačić and Sandra Kovačević
Future Development Plans for Conservation Areas in Taiwan . . . . . . . . . . . . 81

Tung-Ming Lee and Alex Yaning Yen
Innovative Methods on Risk Assessment, Monitoring and Protection

of Cultural Heritage
The Protection of Cultural Heritage Sites from Geo-Hazards:

The PROTHEGO Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Kyriacos Themistocleous, Branka Cuca, Athos Agapiou,
Vasiliki Lysandrou, Marios Tzouvaras, Diofantos G. Hadjimitsis,
Phaedon Kyriakides, Demetris Kouhartsiouk, Claudio Margottini,
Daniele Spizzichino, Francesca Cigna, Giovanni Crosta, Paolo Frattini,
and José Antonio Fernandez Merodo
Sensing the Risk: New Approaches and Technologies for Protection

and Security of Cultural Heritage. The “PRO_CULT” Project. . . . . . . . . . . . 99
N. Masini, F.T. Gizzi, M. Biscione, M. Danese, A. Pecci, M.R. Potenza,
M. Scavone, and M. Sileo
Intangible Cultural Heritage Documentation
The Europeana Sounds Music Information Retrieval Pilot . . . . . . . . . . . . . . 109

Alexander Schindler, Sergiu Gordea, and Harry van Biessum
The Culture of Epigraphy: From Historic Breakthrough to Digital Success . . . 118

Pantelis Nigdelis, Vassilis Bokolas, Nikoleta Vouronikou,
and Pavlos Anagnostoudis
Exploiting Agriculture as an Intangible Cultural Heritage: The Case

of the Farfalla Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Alessandro Pozzebon and Andrea Ciacci
Digital Applications for Materials’ Preservation and Conservation

in Cultural Heritage
Application of Digital Technologies in the Restoration of Historic Buildings

and Heritage Objects: A Selection of Practical Examples . . . . . . . . . . . . . . . 141
Yves Vanhellemont, Michael de Bouw, Liesbeth Dekeyser,
Samuel Dubois, Simon Vermeir, Peter Van Damme,
and Joeri-Aleksander Van der Have
Conservation and Valorization of Heritage Ethnographic Textiles . . . . . . . . . 151

A. Ispas, C. Popescu, G. Roșu, H.C. Rădulescu, H. Fischer, P. Roedel,
M. Dinu, and R. Radvan
Contents – Part II XXVII
Preservation and Valorisation of Morocco’s Ancient Heritage: Volubilis . . . . 160

Muzahim Al-Mukhtar, Ali Chaaba, Mustapha Atki, Rachida Mahjoubi,
Remi Deleplancque, Kévin Beck, Xavier Brunetaud, Romain Janvier,
Khalid Cherkaoui, Issam Aalil, Dalal Badreddine, and Ayda Sakali
Visualisation, VR and AR Methods and Applications
Experiencing Cultural Heritage Sites Using 3D Modeling

for the Visually Impaired . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Kyriacos Themistocleous, Athos Agapiou, and Diofantos G. Hadjimitsis
Multimedia Interactive Map for CH Presentation. . . . . . . . . . . . . . . . . . . . . 178

Nicola Maiellaro and Antonietta Varasano
Interactive Scalable Visualizations of Cultural Heritage for Distance Access . . . 191

Sven Ubik and Jiří Kubišta
Differences of Field Dependent/Independent Gamers on Cultural Heritage

Playing: Preliminary Findings of an Eye–Tracking Study . . . . . . . . . . . . . . . 199
George E. Raptis, Christos A. Fidas, and Nikolaos M. Avouris
The New Era of Museums and Exhibitions: Digital Engagement

and Dissemination
Digital Technologies in the Museum: Same Old, Same Old? . . . . . . . . . . . . 209

Inge Kalle-den Oudsten
A Personal Tour of Cultural Heritage for Deaf Museum Visitors. . . . . . . . . . 214

Vaso Constantinou, Fernando Loizides, and Andri Ioannou
SigNet: A Digital Platform for Hellenistic Sealings and Archives . . . . . . . . . 222

Stefano G. Caneva and Branko F. van Oppen
Places Speaking with Their Own Voices. A Case Study

from the Gra.fo Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Alessandro Pozzebon, Francesca Biliotti, and Silvia Calamai
Paintings Alive: A Virtual Reality-Based Approach for Enhancing

the User Experience of Art Gallery Visitors . . . . . . . . . . . . . . . . . . . . . . . . 240
Stavros Panayiotou and Andreas Lanitis
Digital Documentation and Digital Prototyping of Sacred Artwork

of Museum Museu Mineiro - Brazil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Altino Barbosa Caldeira and Silvio Romero Fonseca Motta
XXVIII Contents – Part II
Parian Marble: A Virtual Multimodal Museum Project. . . . . . . . . . . . . . . . . 256

Marinos Ioannides, Pavlos Chatzigrigoriou, Vasiliki Nikolakopoulou,
Georgios Leventis, Eirini Papageorgiou, Vasilis Athanasiou,
and Christian Sovis
Immersive Digital Heritage Experience with the Use

of Interactive Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Marinos Ioannides, Vasilis Athanasiou, Pavlos Chatzigrigoriou,
Eirini Papageorgiou, Georgios Leventis, Vasiliki Nikolakopoulou,
and Christian Sovis
Digital Cultural Heritage in Education, Learning and Training
Establishing a Remote Sensing Science Center in Cyprus: First Year

of Activities of ATHENA Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Diofantos Hadjimitsis, Athos Agapiou, Vasiliki Lysandrou,
Kyriacos Themistocleous, Branka Cuca, Argyro Nisantzi,
Rosa Lasaponara, Nicola Masini, Marilisa Biscione, Gabriele Nolè,
Ramon Brcic, Daniele Cerra, Michael Eineder, Ursula Gessner,
Thomas Krauss, and Gunter Schreier
The Contribution of Digital Technologies to the Mediation

of the Conservation-Restoration of Cultural Heritage . . . . . . . . . . . . . . . . . . 283
Clément Serain
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

Full Paper: 3D Reconstruction and 3D
Modelling
Implementation and Usage Scenarios of a Participatory
Platform for Cultural Environments
Zois Koukopoulos and Dimitrios Koukopoulos ✉

( )
Department of Cultural Heritage Management and New Technologies,

University of Patras, Agrinio, Greece
(zkoukopu,dkoukopoulos)@upatras.gr
Abstract. Raising audience awareness over the creation and evolution of a

cultural participatory digital platform is a critical point for its acceptance. The
proposed platform adopts user involvement in the content collection level through
the implementation of a mobile application easily downloadable to the user’s
smartphone and the use of a web portal application. Complementary web portal
permits the management of the collected content in a trustworthy manner adopting
an extended role-based access control model for authorization purposes. Users
can formulate private groups to contribute and share content. Platform guarantees
the soundness of contributed content through an auditing procedure requested by
the contributors and conducted by experts selected randomly. In order to stress
the applicability of our platform to various cultural environments, we present a
number of usage scenarios targeting various stakeholders from specialists and
museum curators to students, teachers and simple enthusiasts aiming in the devel‐
opment of coherent narrations.
Keywords: Participatory platform · Cultural content · Web portal · Mobile

application
1 Introduction
Cultural heritage is useful as a means to highlight common origins and achievements

among nations around the world. Cultural heritage management tries to reveal and
present our common origins and cultural achievements, more than any other way [16].
Through cultural heritage management, international structures like the European Union
try to reach the target of Europeanization, the inculcation of the European Idea among
the nations of Europe [6]. The explosion of Internet use around the world provides the
essential background to amateurs and enthusiasts of culture to search for cultural content
or collect and publish their own content [1–3]. This trend poses a series of questions
regarding the quality and the usefulness of such content, in the case it is contributed by
non-experts. Is the contributed content sound or can it be beneficial to specific commun‐
ities? Answering these questions is not a simple task. As the volume of digital cultural
content rises and the technological resources are evolving in terms of efficiency and
availability, the need for cultural data management becomes a priority. New mechanisms
and systems need to be proposed and implemented to manage the produced, vast

M. Ioannides et al. (Eds.): EuroMed 2016, Part I, LNCS 10058, pp. 3–14, 2016.
DOI: 10.1007/978-3-319-48496-9_1
4 Z. Koukopoulos and D. Koukopoulos
amounts of cultural data. Participatory platforms are such systems where data are being
contributed by users in smaller or larger amounts, collected, organized and presented to
the public. Users can range from established scholars and university professors, to simple
enthusiasts of cultural heritage or professionals in the cultural market.
In recent years, several participatory systems are trying to open new ways in cultural
heritage digital management [1–3]. Also there are others that exploit the notion of open
participation, complementary [6]. The first experience from those systems indicates that
not only a participatory system needs to be interactive, dynamic and easy to use, but
also to constantly urge users in involving more and more with the platform and integrate
new services based on participant feedback. Here we propose a participatory system
design oriented to user needs, experts or not, concerning content contribution and
authoring. We involve modern mobile devices in the digitization and content collection
procedures, along with content visualization functionalities through cultural maps. Our
main goal is two-fold: the provision of a learning tool for culture, and a dissemination
tool for expert and public communities to communicate cultural content they create or
collect in a trustworthy manner.
The proposed system design guided the implementation of Culture Gate, a participatory
platform that aspires to become a fundamental digital platform that will receive cultural
heritage content from different sources (scientists, artists, the public etc.), preserve and digi‐
tize it, organize it in widely accepted ways and present it to the online users in a friendly,
modern and innovating manner. By making scientific information public, we allow users to
correct, update or enrich cultural content leading in more qualitative data. Our implementa‐
tion is oriented to satisfy typical usage scenarios concerning cultural content searching,
viewing, uploading and authoring, that may be applicable to various groups of users like
experts, artists, educators, tour operators, public authorities or simple enthusiasts. Culture
Gate takes special care to protect sensitive data (like user credentials and contributed mate‐
rial) through a suitable authorization mechanism based on the extension of role-based access
control model (RBAC model [17]).
The platform offers a dedicated smartphone application which the users can use to
collect cultural content and store it to the system. This capability offers users the oppor‐
tunity to capture tangible or intangible cultural content in real-time and share their
findings with the community in an open or private manner. Field scientists are extremely
benefited by such a service since they could digitize their findings, immediately and
store them in a secure environment for further management.
2 Related Work
Participatory digital platforms implement the participatory action research theory in the
digital domain [8]. Oomen and Aroyo in [18] propose specific classifications of cultural
heritage crowdsourcing platforms: correction and transcription platforms and projects
[7], contextualization [9], complementing collections [10], classification [11] and co-
curation [12]. There are platforms that support content contribution as a key functionality
[1–3], while others support only complementary and occasionally the creation of crowd-
sourcing exhibitions where any user can be a contributor [6]. Moreover, some platforms
Implementation and Usage Scenarios of a Participatory Platform 5
host information about tangible cultural content [1, 3] while others deal with intangible
content [2] or both [6] and platforms that support mobile applications to view and search
content [3, 6], like Muse [3] which hosts native iPad apps for digital libraries with
cultural heritage content. Current trends in cultural heritage management platforms
dictate the integration of mobile services within participatory platforms [1, 3, 9].
MQUADRO offers a dedicated tablet application [1], Muse hosts native iPad apps for
digital libraries with cultural heritage content [3], while an application for android and
iPhone is being used in a portal dedicated to storytelling in Denmark [9].
Security is a critical point concerning the public acceptance of a cultural heritage
management and dissemination system [4]. An important aspect of security in cultural
heritage platforms is the limitation of the user access to sensitive content. An RBAC
based authorization mechanism for secure mobile services used in archaeological exca‐
vations is presented in [19], while an authorization model for musical heritage archives
is described in [20]. A guiding services platform presented in [21] supports trustworthy
content dissemination through an extended RBAC model.
Participatory platforms are gaining ground to market sectors like tourism or social
sectors like education and sociology. TripAdvisor [13] is a successful participatory
platform where users contribute content about sites, locations or businesses that interest
tourists and visitors of a place. The amount of information contributed by users is so
vast and rich that no dedicated project team would ever accomplish the task of collecting
these data. Second Life [14] is a participatory platform that simulates an alternative 3D
world where everyone you see is a real person and all content is produced by users.
Users interact with each other, solve problems and build a collaborative environment,
of mutual learning, achieving the creation a functional society, without supervision of
certified professors or teachers [15].
3 System Architecture and Implementation Technologies
Culture Gate is a general purpose platform that implements various services in order to
support multiple objectives. Platform design addresses several quality issues like scal‐
ability, security, concurrency, integration of third party software, low resource cost and
adaptability to several computational systems (mobile devices or traditional computers),
in a participatory environment.
• Participation: The design should support multiple user roles with different permission
levels that implement a wide range of activities, like content contribution or system
monitoring.
• Scalability: The system needs to be able to support new functionalities and proce‐
dures in order to offer new services or enrich the existing ones, periodically. A static
system is threatened to become out of date very quickly.
• Security: The users that contribute content must be sure that their intellectual property
is protected and editable only by themselves. Also the system must protect the daily
use, from malicious attacks (bots or hackers).
• Concurrency: As a participatory platform, the design should support multiple users
simultaneously.
• Third Party Software Integration: Since the platform wishes to host artwork from
artists or experts, the design must be able to support efficiently various software
implementations with different technologies.
• Low Cost: The design must use limited resources in equipment, software and human
dynamic to become viable and extend its lifecycle.
• Adaptability: The platform should be accessible both from traditional computational
systems (desktops, laptop) and mobile devices (smartphones, tablets).
In order to support all the above design goals, the platform uses a modular architecture
consisting of several modules (Fig. 1):
Fig. 1. Culture Gate system architecture
Culture Gate Mobile Application (CGMA). Smartphone application that collects

cultural content from indoor or outdoor locations, in real-time and stores it in platform
database and file system. CGMA feeds the system with cultural data.
Culture Gate Online Web Portal (CGOWP). Web interface where users can submit
content or search and view cultural information. CGOWP presents the cultural content
to the broad public.
Web Server. Supports the necessary tools for multiple operations like request handling
from CGMA and CGOWP, hosting and managing system database and file system.
Request Handler. Implements scripts that get data from CGMA and CGOWP content
contributors perform insert/update/search/delete operations to system database, create
responses to send encoded data to CGMA users and displays user requested content.
File System. All files are stored in the platform file system. The files vary from php,
html, xml scripts and apache files to directories with audiovisual material from users’
contributions (CGMA and CGOWP). This system is isolated from all the other modules,
except the handler, which is the only module that has permission to interact with the file
system, for security reasons.
System Database. Stores all information concerning application data (including user
credentials) and cultural data, contributed by the users.
The system follows RBAC model in order to classify users to several roles/groups
(administrator, moderator, auditor, contributor, guest). Each user can have more than
one role. Furthermore, platform users can be registered users (moderator, auditor,
contributor), simple users (guest) or members of the project team (administrator, moder‐
ator). In Table 1, we present user roles, types and their permitted operations.
Table 1. User groups/roles.

Role Type Operations
Administrator Project team Supervises every technical procedure, checks
system integrity, assigns moderator roles to
users, administers system database and file
system, views all content (published and
unpublished), deletes users
Moderator Project team – registered Assigns user roles, creates user private groups,
user views all content, checks for offensive content
and warns or bans users for this purpose
Auditor Registered user Views private content after assignment, informs
contributors for the required content
modifications
Contributor Registered user Views public content and private content of his
group, uploads content. A contributor can modify
only the content he contributes
Guest Unregistered user A nomadic user who views and searches public
content
Platform supports different types of content (text, audio, images, video and multi‐
media). Text content can be content description, notes, event announcements, historical
background information, theatrical plays and scripts, literary works or reviews and
references. Images can be site and artifact pictures or depictions (like drawings, pictures,
notes, sketch). Video can be animation or live activities like excavation or a folk dance.
Audio content can be narrations, storytelling, interviews and vocal guides. Multimedia
files can be a full demo package with all kinds of data types.
The basic entity of contributed content is the cultural item which is characterized by
three categories of attributes: content-based, context-based and model-based attributes.
Content-based attributes include thematic ones like title and description, and disciplines
where the contribution belongs to. Context-based attributes include featured image
(contribution’s most representative image), files (multimedia files corresponding to the
contribution), location (region’s geographical coordinates where the contribution was
uploaded), digitization equipment (equipment used to collect the data) and date (date
and time of the contribution). Model-based attributes include the permitted operations
on the specific item by specific user roles (like authoring or content visibility).
Culture Gate platform uses several technologies to implement all its modules:
Apache Web Server, PHP scripts, MySQL database, Google Maps API and Android
operating system (JAVA, JSON, XML). For the implementation of CGOWP we used a
Windows 10, intel core i7-2600, 8 GB RAM at 1866 MHz system. For CGMA imple‐
mentation we used Android Studio as the IDE, target SDK version 23 and minimum
SDK version 8 and Samsung Galaxy Grand Neo to test the resulted application (Fig. 2).
Fig. 2. Cultural listings table schema (MySQL workbench)
4 Platform Services
Platform services are classified in three categories according to the sub-system they refer
to: core, mobile and web portal services.
Core Services. Security: We propose an extended RBAC model (implemented by

Request Handler) for authorization purposes. Platform supports five user roles with
different authorization privileges to platform’s content (administrator, moderator,
auditor, contributor and guest). User passwords are encrypted (PKI) and disclosed only
to the user. Several programming techniques are applied to encounter security breaches.
Coding techniques used for security reasons are: input validation, compiler warnings
heeding, simple security design, default denying, sanitizing data sent to other systems
and effective assurance techniques. Registration: Creation of a new user account. Users
complete fields like username, email, password and discipline. Moderation: The plat‐
form allows registered users with elevated permissions to monitor the daily procedures
and keep the platform live (like monitoring disciplines, CGMA monitoring, cultural
information uploading and management). Authoring: Guarantees that platform content
is sound and up-to-date. After assignment to a specific contribution by the moderator,
the auditor checks the validity of the contributed content and informs the contributor for
the needed changes. Group formation: When a registered user wants to formulate a
private group to share content, in a protected manner, with other users, he communicates
with a moderator. This user must provide a clear group description (like a class of
students that implements an educational virtual tour to a museum) and a list of platform
users who will serve as group members. This user will be the group leader and will be
responsible for the content each member of the group uploads. This content will be
private and restricted only to the group members for viewing. A group member that
uploads private content must declare that this content is only available for the group
members. A group of colleagues could create a private user group to share real-time
content with each other. Personalization: The platform displays to a user, information
according to his interests. User chooses fields of interest during registration procedure.
Mobile Services. CGMA’s main task is to collect and upload digitized cultural content.
CGMA offers three different services that can be used by several user groups to store
real-time data. Field service: Capturing audiovisual content from an outdoor uncharted
location (like an excavation point) and sending it with its coordinates and notes to the
web server. Users complete fields like: Title, Description, Discipline, Audio/Video/
Image Files. CGMA sends geographical data automatically during the uploading proce‐
dure. Street service: Capturing audiovisual content from an outdoor charted location
(like a city street) and sending it with its coordinates and notes to the web server. Indoor
service: Sending text content related to a tangible cultural item (like an artifact in a
museum), along with notes or audiovisual files to the web server.
Web Portal Services. CGOWP supports management, presentation and dissemination

services. Content organization: When a user uploads content from data submission
screens, he makes a series of choices that give semantics to each new record (like disci‐
pline). User has the ability to view related content to a specific point in cultural map.
Content viewing: A user has two choices in order to view content: geographical map or
list. Content searching: A user has several choices in order to search content: keywords,
disciplines and geo-location data or their combinations. Dissemination and User attrac‐
tion: calendar of events, news, serious games implementation, 3D representations,
smartphone applications and time-lapse videos.
5 Usage Scenarios
Culture Gate is a collector, organizer and presenter of cultural heritage content. In order
to highlight the platform capabilities, we present specific scenarios based on the used
service carrier: web portal or mobile device.
Web Portal Based Scenarios
Guest Searching and Viewing. User accesses a cultural map where he views all content
as pins on a geographical map and lists of cultural information, organized per discipline.
The map displays every public listing, contributed by registered users, as a pin. Pins are
grouped by discipline and can be filtered by keywords or disciplines (one or more).
Every discipline has a different color and icon for its pin. Figure 3 displays the cultural
map with all information and Fig. 4 shows a filtered version of the map with the keyword
“Acropolis” and the discipline “Museology”. When the user clicks on a pin, an infor‐
mation box is displayed with some information. If the user clicks on the information he
will be transferred to listing’s dedicated webpage where he can find all the needed
information (Fig. 5).
Fig. 3. Unfiltered view of cultural map
Fig. 4. Filtered view of map with keyword Fig. 5. “Acropolis Museum” listing page
“Acropolis” and discipline “Museology”
Registration. Every user can register to Culture Gate by accessing the “User Registra‐
tion” screen. The user submits credentials (username, password) and information about
the discipline he serves mandatory. Optionally, user declares specialty and fields of
interest. This procedure is vital for system, to personalize the presented content
according to user interests and draw information to formulate the auditor pool.
Authoring. When a registered user makes a content contribution and marks it as “Needs
Authoring” then the system initializes the authoring procedure, which is a comprehen‐
sive way to check the soundness of platform’s content. After contributor request, the
content moderator chooses one of the auditors with related specialty randomly to audit
the content. If the auditor accepts the invitation, he proceeds to the submitted content
checking. Afterwards, the auditor sends the content back to the contributor with the
proposed modifications. When the contributor makes the appropriate modifications, the
auditor gives his approval for the content to be published as audited. If an auditor rejects
invitation, another auditor is chosen.
Moderation. Culture Gate uses a decentralized model to administer platform content

and everyday use. Administrator chooses the users that will become moderators from a
pool of registered users that have shown interest for this role. According to the assigned
duties (like monitoring disciplines or contributed content), the administrator gives the
corresponding elevated permissions. For example, if a user is chosen to become a content
moderator, the administrator will allow him to moderate contributions, create private
user groups or ban users, but he will not be able to moderate content assigned to another
moderator.
Registered User Public Content Uploading and Managing. A registered user is

permitted to upload cultural content through the “Add Listing” screen. Users contribute
content filling out fields like title, description, discipline, featured image, multimedia
files, location, date or keywords. Users can click the “Review your Listing” button and
see the way the content will be displayed. If the content is compliant to platform’s terms
and conditions, the information is being published. When the uploading procedure
terminates, the content becomes available to the public (Fig. 3).
Registered User Private Content Uploading and Viewing. A registered user that
uploads content has full ownership over the content. The user can declare his content
public (for everyone to see) or private (can be seen by him or a certain user group). When
a registered user wants to formulate a private group to share content, in a protected
manner, with other users, he informs a moderator. The user must provide a clear group
description (like a class of students that implements an educational virtual tour to a
museum) and a list of users who will serve as group members. The first user will be the
group leader and will be responsible for the content each member of the group uploads.
This content will be private and restricted only to the group members for viewing. A
group member that uploads private content must declare that this content is only avail‐
able for the group members.
Mobile Based Scenarios. CGMA offers an extremely useful functionality to scientists

of cultural heritage, especially those working in the field. These experts can capture
tangible and intangible cultural content and store it immediately and easily to a web
server. This procedure gives them the opportunity to secure their findings, digitize
content and preserve it, share their content with colleagues in real-time and exchange
information with them. Every new finding, is enriched with several other information,
automatically, like the coordinates of the location that the discovery have been made
and annotations about the finding that enrich the content of the discovery.
Field Use - Archaeological Excavation Site. A team of archaeologists works on an

uncharted area. At least one member of the team has an android smartphone with CGMA
installed in it. This smartphone must have an internet connection when CGMA is being
used. The user clicks and opens the application. If he has already used the application
or has an account, he can sign in immediately. Otherwise he must register using the
“Sign Up” screen (Fig. 6) submitting username, password, email and discipline. After
signing in the system, the user navigates to “Field Functionality” screen (Fig. 7) and
submits contribution title, description and the concerning discipline (in this case
“Archaeology”). User can take a photo, record video or audio. When the user clicks the
“Send Cultural Data” button, the system sends the collected data along with the exca‐
vation point’s geographical coordinates to the web server.
Fig. 6. CGMA “Sign Up” screen Fig. 7. CGMA “Field Functionality” screen
Indoor Use - Museum School Visit. A class teacher decides to visit the city museum
with his students and communicates with a moderator in order to create a private user
group for the class, with him as a group leader. Moreover, he asks from every student
to install CGMA to their smartphones and create accounts. The teacher informs the
moderator about the user accounts to be added to the user group. When the class visits
the museum, the teacher informs the students to sign in Culture Gate and navigate to
“Indoor Functionality” screen and start using CGMA’s indoor service. He consults them
to capture artifacts, make notes stating their point of view and send them to Culture Gate.
The teacher can use CGMA to record an audio file with museum’s conducted guide.
When the visit is over, the teacher accesses the content contributed by students and
reviews it. Then, he publishes the content with visibility only to group members. In the
next lesson, he will be able to discuss this museum visit with the students and publish
the corresponding virtual tour to the public.
Street Use - Cultural Event. A user is watching a religious ceremony in a city street.
The user signs in the system and navigates to the “Street Functionality” screen. In order
to capture a video he must submit title, description, discipline and click the corre‐
sponding button to capture a video. After the video is recorded and stored to the smart‐
phone, it is ready to be transferred along with geographical coordinates to Culture Gate
web server by clicking the “Send Cultural Data” button (Fig. 8).
Fig. 8. Religious ceremony in Arta, Greece – video capture

6 Discussion and Future Work
In this work we address the problem of provisioning digital tools that support public
participation in evolving cultural projects. Digital tools are implemented through a
participatory platform where users collect and disseminate cultural information. Since
May 2016 the platform is accessible for the broad public [5]. Web portal services
(content organization, viewing, searching, dissemination and user attraction) have been
well - received by the rapidly growing Culture Gate Community. Most of platform’s
core services (access control, registration, moderation, authoring, group formation,
personalization) have been successfully implemented and tested, except issues related
to intellectual property, which are in design phase. Concerning CGMA, we distributed
the application to a group of application developers to express their opinion about several
technical issues like network connectivity or big data uploading. Simultaneously,
CGMA was distributed to cultural experts to notify us about the application’s usability
and user friendliness. We plan specific activities for elementary and secondary education
to test the acceptance of the proposed platform as an education tool. In this direction we
have already come in contact with school teachers. Those activities will give us feedback
for the improvement of our system.
Alongside with the platform development strategy, we follow an operational strategy
with the formulation of project teams, which consist primarily of interested cultural heritage
academic students, which undertake daily tasks and help in maintaining the platform.
Culture Gate could be commercially exploited in many ways as it can have a significant
social impact in sectors like tourism or cultural heritage market. Museum curators can
promote museum collections by uploading information about them to the platform. Tour
operators or Public authorities could use Culture Gate to highlight certain cultural heritage
spots and attract new clients and tourists, thus helping local and national markets. Organ‐
izers of cultural events have the opportunity to advertise their events to an immense and
targeted public. Moreover, CGMA could contribute revenue in two different ways: the free
version could host advertisements while a premium version with additional features could
become available for downloading for a small price. In order to handle intellectual property
rights we will enhance our platform with watermarking services for image, audio and video
that will be available with a small fee to any contributor upon the uploading of his contri‐
bution. For images, we plan to offer a service based on a watermarking scheme that guar‐
antees efficient and fast embedding and detection of watermarks [22]. For MP3 audio files,
we will provide a service based on an algorithm that supports watermarks with semantic
meaning and it is fast and reliable for online use [23]. For video files, the selection of an
efficient scheme will be a goal for future work.
References
1. Lacedelli, S.Z., Pompanin, G.: MQUADRO: a platform model for cultural heritage. In: 104th
Annual Conference of the College Art Association, Washington (2016)
2. The Prow. http://www.theprow.org.nz/
3. Muse Platform. http://muse-opensource.org/
4. Arnold, D.: Pasteur’s quadrant: cultural heritage as inspiration for basic research. ACM J.
Comput. Cult. Heritage 1(1), 1–10 (2008)
5. Culture Gate Web Portal. http://www.culture-gate.com
6. EUROPEANA. http://strategy2020.europeana.eu/
7. Transcribe Bentham Project. http://blogs.ucl.ac.uk/transcribe-bentham/
8. Rahman, M.A.: People’s Self-development: Perspectives on Participatory Action Research:
A Journey Through Experience. Zed Books, London (1993)
9. Stories of Denmark. http://www.kulturarv.dk/1001fortaellinger/en_GB
10. UK_Soundmap Project. http://sounds.bl.uk/sound-maps/uk-soundmap
11. Flickr: The Commons. https://www.flickr.com/commons
12. Click! A Crowd-Curated Exhibition. https://www.brooklynmuseum.org/exhibitions/click/
13. TripAdvisor. http://www.tripadvisor.com
14. Second Life. http://secondlife.com/
15. Ondrejka, C.: Education unleashed: participatory culture, education, and innovation in second
life. In: Salen, K. (ed.) The Ecology of Games: Connecting Youth, Games, and Learning, pp.
229–252. MIT Press, Cambridge (2008)
16. Michigan State University: Department of Anthropology Cultural Heritage Informatics
Initiative (2010). http://chi.anthropology.msu.edu/program/
17. Ferraiolo, D., Kuhn, R.: Role-based access controls. In: 15th NIST-NCNC National Computer
Security Conference, Baltimore, pp. 554–563 (1992)
18. Oomen, J., Aroyo, L.: Crowdsourcing in the cultural heritage domain: opportunities and
challenges. In: 5th International Conference on Communities and Technologies, pp. 138–149.
ACM Press, Brisbane (2011)
19. Koukopoulos, D., Tsolis, D., Gazis, M., Skoulikari, A.I.: Secure mobile services for on-going
archaeological excavations management and dissemination. In: 6th International Conference
of Information, Intelligence, Systems and Applications, pp. 1–6. IEEE Press, Kerkyra (2015)
20. Koukopoulos, D., Tsolis, D., Heliades, G.: Ionian music archive: application of digitization,
management and dissemination technologies for musical cultural heritage. In: 5th
International Conference of Information, Intelligence, Systems and Applications, pp. 239–
244. IEEE Press, Chania (2014)
21. Koukopoulos, D., Styliaras, G.: Design of trustworthy smartphone-based multimedia services
in cultural environments. Electron. Commer. Res. 13(2), 129–150 (2013)
22. Tsolis, D., Sioutas, S., Papatheodorou, T.: A multimedia application for watermarking digital
images based on a content based image retrieval technique. Multimedia Tools Appl. 47(3),
581–597 (2010)
23. Koukopoulos, D., Stamatiou, Y.: An efficient watermarking method for MP3 audio files. In:
International Enformatika Conference, pp. 154–159. Enformatika, Prague (2005)
Benchmarking Close-range Structure from
Motion 3D Reconstruction Software Under
Varying Capturing Conditions
Ivan Nikolov(B) and Claus Madsen
Architecture, Design and Media Technology, Aalborg University,

Rendsburggade 14, 9000 Aalborg, Denmark
{iani,cbm}@create.aau.dk
http://www.aau.dk/
Abstract. Structure from Motion 3D reconstruction has become widely

used in recent years in a number of fields such as industrial surface
inspection, archeology, cultural heritage preservation and geomapping.
A number of software solutions have been released using variations of
this technique. In this paper we analyse the state of the art of these soft-
ware applications, by comparing the resultant 3D meshes qualitatively
and quantitatively. We propose a number of testing scenarios using differ-
ent lighting conditions, camera positions and image acquisition methods
for the best in-depth analysis and discuss the results, the overall perfor-
mance and the problems present in each software. We employ distance
and roughness metrics for evaluating the final reconstruction results.
Keywords: Multi-view 3D reconstruction · Structure from Motion

(SfM) · Photogrammetry · Software comparison · Benchmark
1 Introduction
Structure from Motion (SfM) for 3D reconstruction has come a long way in recent
years. The technology is at a point where a multitude of commercial and free
packages exist, enabling non-experts to quickly and easily capture high quality
models from uncalibrated images. An example is given in Fig. 1.
Most of these packages are used for landscape reconstruction, creation of
orthomosaics and large scale reconstructions. They can be also used for close-
range reconstructions. This makes them perfect for use in cultural heritage
preservation, artifact digitalization, virtual museums and others. However, many
of these solutions come with high initial and upkeep monetary costs. This makes
choosing the one most suitable for a specific task an important first step for
each project relying on 3D reconstruction - both in result accuracy, resource
requirements and performance across varying conditions. Such an endeavour
can require a large investment of time. This is why in our paper we provide
an in-depth overview of the newest and most widely used commercial software

c Springer International Publishing AG 2016
DOI: 10.1007/978-3-319-48496-9 2
16 I. Nikolov and C. Madsen
Fig. 1. Example of 4 out of N input images, taken from different view points and the
resulting camera position triangulation and dense point cloud creation. The view is
from Agisoft PhotoScan.
solution tested across various conditions. We concentrate on close-range SfM, as

opposed to aerial or long-range.
Six commercial 3D reconstruction software solutions are chosen for testing
in the paper. Each of the solutions takes an unordered list of images as input,
extracts features and creates a sparse point cloud, triangulating the camera
positions. A dense point cloud and a mesh are created by interpolating the
sparse point cloud. Texture of the reconstructed object is also created.
Six different objects are used for the reconstructions, depicted in Fig. 2. They
are selected according to their varying reconstruction difficulty and different
problems that they present like textureless surfaces, repeatable patterns, sym-
metrical objects, glossiness, etc. Objects are scanned with a white light scanner
for evaluating the meshes produced by the SfM packages.
Six different scenarios are tested. These scenarios cover different lighting,
positioning and shooting setups. These experiments show that the environmental
conditions have a noticeable impact on the final reconstruction and affect some
software solutions more than others.
For verifying the accuracy of the output meshes from the different programs,
two qualitative methods are chosen: (1) calculating the signed distance between
ground truth objects and the reconstructions; (2) comparing the local roughness
profiles between the ground truth objects and the reconstructions. The results
show that some of the tested packages have more problems reconstructing glossy,
symmetrical and textureless surfaces, than others, resulting in complete failures.
Some programs sacrifice details for a less noisy final mesh, while others capture
more detail, but are very sensitive to noise. A moving camera setup with uniform
lighting also gives higher reconstruction accuracy than a turntable setup.
2 Related Work
SfM is just one of many techniques for 3D reconstruction of objects and artifacts.
Other techniques are beyond the scope of the paper, but for a quick overview the
work in stereo-vision reconstruction [13], structured light [8] or laser scanning
[14] is available for reference.
For SfM reconstruction most resources for benchmarks and comparisons are
either from archaeological context [15] or from geomapping context [16]. These
give valuable information, but are mostly focused on one type of surfaces and
Benchmarking Structure from Motion 3D Reconstruction Software 17
Table 1. Tested software solutions with some of their most important characteristics.
In the output column the shortened names denote: dense point cloud (DPC), sparse
point cloud (SPC). The price is given for both standard and pro versions. Bolded font
denotes the one used for testing.
Program Outputs Online/Offline OS Scripting Price (USD)

ContextCapture Mesh/Texture/DPC Offline Win Yes N/A
Memento Mesh/Texture Online/Offline Win/Mac No Free/190 annual
PhotoScan Mesh/Texture/SPC/DPC Offline Win/Mac/Linux Yes 179/3499
Reality Capture Mesh/Texture/SPC/DPC Offline Win Yes 110/8351 annual
3DF Zephyr Mesh/Texture/SPC/DPC Offline Win No 199/3200
Pix4D Mesh/Texture/DPC Offline Win/Mac No 3222 annual
objects to reconstruct under a more limited set of environment conditions. Other

resources [9,17] give more in-depth comparison using both their own datasets
and freely available ones, but lack the comparison for a larger number of software
solutions.
3 Tested Software
We have chosen six of the state of the art software packages for 3D reconstruc-
tion. These products are Agisoft PhotoScan Pro [3], Bentley ContextCapture [1],
Autodesk Memento (ReMake) [2], Pix4D [5], 3Dflow 3DF Zephyr Pro [6] and
Reality Capture [4]. For more information on some of the important features
each of the selected software solutions has, please refer to Table 1. The prices are
subject to change and are given as they are in the time of writing this paper and
converted to dollars. In the output column four of the most widely used ones
for close-range photogrammetry are given to preserve space - 3D mesh, tex-
ture, sparse point cloud, dense point cloud. Additional outputs like orthophotos,
orthomosaic, fly-through videos, depth and normal maps, etc. are supported by
many of the programs, but are out of the scope of this paper.
4 Datasets
The six chosen objects are shown in Fig. 2. These objects are selected based on a
number of criteria concerning the properties of the materials that they are made
of. These criteria are used to judge the capability of each software to handle dif-
ferent difficult cases, which are considered weak points for SfM. The criteria are
as follows - glossy/smooth surfaces, monochrome colors, very dark/black color,
repeating patterns, partial occlusions, symmetrical surfaces. They may result in
failures in reconstruction, decreased overall accuracy, cause holes and noise in
the resultant point clouds and mesh [20,21].
As an initial observation the objects are divided into two groups depending
on their perceived reconstruction difficulty. The easy to reconstruct objects -
angel statue, sea vase and bird bath and the hard to reconstruct objects - black
vase, plastic vase and owl statue. The angel statue and sea vase are perceived
as easy because they have a lot of surface detail and features, both global and
local, which should make them easy to reconstruct by all the programs. They
also have some partial self occlusion, which will be tested. The bird bath is also
feature rich and has both very smooth and glossy surface parts, as well as rough
ones. The black and plastic vases are perceived as hard, because of their color,
glossiness and repeated patterns. The owl statue is chosen as an intermediate
object, which has a lot of glossiness and feature poor parts, as well as non-glossy
more feature-rich ones.
Fig. 2. Testing objects: (a) angel statue, (b) bird bath, (c) sea vase, (d) plastic vase,
(e) owl statue, (f) black vase. Typical size of the objects is between 25 and 60 cm
The input images are taken using a Canon 6D camera at maximum res-
olution of 5472 × 3648. A zoom lens with focal length of 70–300 mm is used
to accommodate the different zoom levels needed for the different objects. The
reconstructions are carried out on a stand alone laptop equipped with Intel Core
i7 - 4710HQ at 2.50 GHz, 16 Gb of RAM and a GeForce GTX 970M. The oper-
ating system is Windows 8.1. Each of the six objects has been scanned with a
high resolution, high accuracy white light scanner from Aicon. These scans are
considered detailed enough to be used as a benchmark for the performance. To
demonstrate the accuracy and detail of the scans, a cube with known dimensions
is also scanned and the measurement of the 3D model’s sides are compared to the
real world ones. The two differ by an average of 1.03 mm/1.12 mm/0.93 mm in
width/height/depth. Henceforth these scans will be referred to as ground truth
objects, while the outputs from each of the tested programs will be referred to
as reconstructed objects.
5 Testing Scenarios and Results

5.1 Main Test Scenario
All six objects are used in the initial test scenario, together with all the tested
programs. The distance between the reconstructed and ground truth objects is
calculated, together with the local roughness. The scenario aims to determine
how are the selected programs fairing when tested with both easy and hard
to reconstruct objects, as well as noting their speed, accuracy and robustness
against noise. The test also aims to determine the object factors which make
reconstruction hardest for each of the programs.
The test scenario uses photos captured in an indoor controlled environment.
The image capturing algorithm is as follows - the captured object is positioned
on a turntable; the camera on a tripod is facing the object and is lower than it
for capturing the first set of images at lower angle; one light is positioned on a
stand above the camera so it shines directly at object; a photo is taken and the
turntable is turned 20o ; this is done 18 times, so the object is captured from all
sides; the camera on the tripod is then repositioned higher two times, each time
18 more photos are taken; a total of 54 photos of the three different height sets.
The CanonD6 camera is used for taking photos as it gives high detail photos,
without straining the hardware of the testing machine.
The total processing time of creating the 3D model is noted for each program.
For the online version of Memento, the processing time does not give a proper
estimate of the working time. It is given just for a more full presentation of the
data. This data is given in Table 2. A course visual inspection is done on the
created model, focusing on severe problems with the objects.
Table 2. Processing time in seconds for each of the six objects by the tested software
solutions. Models which contain problems like missing sides, broken parts, floating
noise, etc. are marked with red. Models which could not be reconstructed are given an
N /A notation.
Program Angel Bird bath Owl Sea vase Plastic vase Black vase
ContextCapture 2820 s 3600 s N/A 3780 s 3060 s N/A
Memento Online 4860 s 4920 s 5160 s 4440 s 4260 s 5340 s
PhotoScan 4020 s 4500 s 3780 s 4560 s 4740 s 3480 s
RealityCapture 5220 s 6480 s N/A 6720 s 2820 s N/A
3DF Zephyr 3720 s 4440 s 4140 s 4860 s 3060 s 4680 s
Pix4D 4140 s 3240 s N/A 4860 s 3960 s 3720 s
Memento Offline 11520 s 9360 s 7140 s 10320 s 7980 s N/A
All packages, except Memento offline have comparable processing times,

which depend on the complexity of the reconstructed object. Memento online,
PhotoScan and 3DF Zephyr could reconstruct all objects, while ContextCapture,
Reality Capture and Pix4D experienced the most problems. The coarse visual
inspection is followed by a more qualitative inspection, using the ground truth
scanner data for comparison. The idea suggested by Schning and Heidemann
[17] is used for this part of the test scenario. In their paper they conclude that
each tested program produces 3D models and point clouds of different density,
which also may contain parts of the background or noise particles. Therefore, it
is better to use the reconstructed models as reference and compare the ground
truth data to each, noting the difference. In addition their idea of using the
meshes for comparison is used, as opposed to using the point cloud. This gives
the possibility to test signed distances using the model’s pre-calculated normals.
The comparison between the reconstruction and the scanned data is done
using the free open source software CloudCompare [7]. The reconstructed models
are scaled to the absolute scale of the ground truth and registered to it using
an iterative closest point algorithm (ICP) by Besl and McKay [18]. Once the
models are registered the distances between the triangles of the reconstructions
and the ground truth is calculated. Using the normals of the meshes the distance
is calculated as signed. These distances are visualized as pseudo color heat map.
The pseudo color maps for the angel statue can be seen in Fig. 3. The maps
are filtered removing distances outside the interval of [−0.3 mm; 0.3 mm], for
easier visualization. From the distances, the mean and standard deviation of the
distance distribution for the whole object are calculated. A Gaussian normal
distribution is assumed for the modelling of the distance distribution between
the ground truth and the reconstruction. The mean and standard deviation are
given in Table 3 for the easy to reconstruct objects - angel statue, bird bath and
sea vase, together with the Gaussian distributions for them in Fig. 4. For the
hard to reconstruct objects - the plastic vase, owl and black vase the data is
given in Table 4 and Fig. 5, respectively.
Fig. 3. Pseudo color distance maps between the ground truth and the reconstructed
objects. Red colors indicate distances above the ground truth, blue colors indicate
distances below the ground truth and green colors indicate where the surfaces coincide.
(Color figure online)
Table 3. Mean value (μ) in mm and standard deviation (σ) in mm2 of the distance
metric for each software solution for the three objects selected as easy to reconstruct
Angel Bird bath Sea vase

Mean/Variance Mean/Variance Mean/Variance
ContextCapture −0.024/0.703 −0.030/0.588 −0.245/2.016
Memento Online −0.089/0.438 −0.039/0.382 −0.408/2.277
PhotoScan −0.109/0.805 0.034/0.175 −0.463/2.321
Reality Capture −0.038/0.486 −0.006/0.143 −0.481/2.421
3DF Zephyr −0.040/1.020 −0.045/1.537 −0.911/3.514
Pix4D −0.194/1.124 −0.060/0.668 −0.425/2.419
Memento Offline −0.080/0.569 −0.046/0.40 −0.255/2.983
Fig. 4. Histograms of the Gaussian distribution characterizing the distances between

the ground truths and the three easy objects. All the histograms are scaled the same.
metric for each software solution for the three objects selected as hard to reconstruct
Plastic vase Owl Black vase

Mean/Variance Mean/Variance Mean/Variance
ContextCapture −2.512/10.601 N/A N/A
Memento Online −3.450/6.697 −0.937/3.318 −4.549/5.886
PhotoScan −3.791/7.027 0.371/6.806 −4.331/5.758
Reality Capture −4.395/7.222 N/A N/A
3DF Zephyr −4.814/7.471 0.169/3.191 −4.035/5.933
Pix4D −3.782/7.187 N/A −4.794/6.027
Memento Offline −5.074/7.429 −0.929/0.977 N/A
Fig. 5. Histograms of the Gaussian distribution characterizing the distances between

the ground truths and the three hard objects. The missing histograms are programs
which failed to reconstruct the object. All the histograms are scaled the same.
The initial speculation dividing the objects into easy and hard ones is proven
by the amount of reconstruction failures. Both the black vase and the owl statue,
experience much higher number of failures, compared to the other objects. The
plastic vase fairs better, but because of its symmetrical featureless and dark
surface, the reconstruction suffers from improperly placed geometry. This can
also be seen from the Gaussian histogram distributions in Fig. 5, where the
distributions for both the black vase and the plastic vase are much wider, showing
larger divination distances from the ground truth. The owl statue has less noisy
histogram, but it suffers from holes in the reconstruction. ContextCapture and
Reality Capture demonstrate the overall smallest mean and variance deviations
from the ground truth for the easy objects, but both programs completely or
partially fail when the surfaces are not optimal. 3DF Zephyr, Memento Online
and PhotoScan on the other hand are much more consistent and have a more
graceful degradation of performance, but tend to miss smaller details and have an
overall high variance in the distance distribution. From here another observation
can be made - the programs can be roughly divided into ones that capture a
lot of small detail at the price of noise and easier failures - Context Capture,
Reality Capture, Memento Offline and the ones that are more consistent and
robust, but fail to capture details - Memento Online, 3DF Zephyr, PhotoScan.
Pix4D is mainly aimed at aerial photos and this clearly shows, as the program
is much noisier in all instances.
To determine the amount of noise and over-smoothing of features in the
reconstructions compared to the ground truth, a second metric is introduced.
The local roughness of both the reconstructions and the ground truth is cal-
culated using the Gaussian curvature of the models, which is normalized to
give proper weights to rough patches and smooth patches near edges. The
method is introduced by Wang et al. [19] for assessment of mesh visual quality.
Fig. 6. Pseudo color roughness maps of the ground truth and reconstructed meshes.
The colors go from red to blue through green, depending on how rough the surface is.
The method is useful in the case of our paper as it generates an accurate rough-
ness map, which can give both visual and more quantitative information for the
success of the reconstruction. The roughness map is also visualized as a pseudo
color map, which is given in Fig. 6.
From the local roughness map, the histograms of both the ground truth and
the reconstructions is calculated. Using these histograms the Kullback-Leibler
distances [22] between the ground truth and reconstructions are calculated. This
gives a measurement of the similarity between the two, which penalizes devia-
tions from the roughness of the ground truth both caused by introduction of
noise in the reconstructed mesh and in over-smoothing details in it. Figure 7 has
the results from the roughness histogram distances, where smaller values give
more faithful reconstructions, roughness-wise.
The results from the roughness metric support the division of the programs.
Pix4D introduces a lot of noise and smooths details. This can be seen in both
Figs. 6 and 7, where it has clear disadvantage in many of the cases. Figure 6
also shows that Pix4D, Memento Offline and Reality Capture have introduced a
lot of noise on the smoother parts of the angel, compared to the ground truth,
like the stomach and legs, while Memento Online, 3DF Zephyr and PhotoScan
have smoothed out small features in the face and hair of the angel. Memento
Offline, Reality Capture and PhotoScan manage to capture most of the detail
on the easier to reconstruct objects like the angel and the bird bath, without
introducing too much noise as evidenced by the smaller histogram distances.
However they fail on the smoother objects like the plastic vase and the owl,
where they introduce uncertainty noise. Memento Online and 3DF Zephyr tend
to over-smooth the surfaces as evidenced by the bar chart of the sea vase.
Angel Bird Bath Sea Vase
Plastic Vase Owl Black Vase
Fig. 7. Bar chart visualizing the calculated Kullback-Leibler distances between the
roughness histogram of the ground truth and the reconstructed objects. The tested
software is denoted with short names - Context Capture (CC), Memento Online (Mon),
PhotoScan(PS), Reality Capture (RC), 3DF Zephyr (3df ), Pix4d (Pix) and Memento
Offline (Moff ).
5.2 Follow Up Test Scenarios

One of the best performing objects - the bird bath is tested in a number of
follow up scenarios under different capturing conditions. This is done to deter-
mine the effect of capturing conditions on the reconstruction results. Five follow
up experiments are carried out focusing on different combinations of conditions.
First tested condition is the effect of rotating the camera to capture images from
different views, as opposed to using a turntable to rotate the object and keep
the camera stationary. This test aims to assess if a moving background and com-
pletely static lighting will help with reconstruction process, as opposed to the
lighting which “moved” with the object in the case of using a turntable. The
second tested condition is using multiple light sources for a more even lighting,
as opposed to one directional light. The third condition is using different num-
ber of photo positions, combined into bands of photos, with varying height. Five
and three bands of photos are created. The first two contain 18 photos each
taken in 20o intervals, the next two contain 9 photos in 40o and the final one
contain 4 images. The setup also aims to test if introducing information from
more angles can help the feature point matching algorithm of the tested soft-
ware solutions. The same analysis pipeline is used as with the main experiment,
using the ground truth scans to compare with. The results from the different
combination of conditions are given in Table 5.
The tests show that using static lighting and moving background without a
turntable yields a higher accuracy, with lower mean and standard deviation val-
ues, compared to the turntable results. There is also a difference between using
multiple light sources and just one directional one, with the latter introducing
more noise, which can be seen by the higher standard deviation in the table
metric for the bird bath object for each of the tested software solution from the five
tested shooting scenarios.
No turntable Turntable
Multiple lights One light Multiple lights
Five bands Three bands Five bands Five bands Three bands
(μ)/(σ) (μ)/(σ) (μ)/(σ) (μ)/(σ) (μ)/(σ)
ContextCapture −7.167/13.289 −4.147/8.052 N/A N/A N/A
Memento −0.209/2.028 −0.094/1.306 −0.366/2.148 −0.947/3.826 −0.309/2.305
PhotoScan −0.283/2.312 −0.240/2.685 0.206/1.982 −0.212/2.410 −0.167/1.159
Reality Capture −0.031/0.284 −0.014/0.689 N/A 0.108/1.710 N/A
3DF Zephyr −0.039/0.584 0.011/0.411 −0.308/2.035 −0.372/2.712 −0.165/0.922
Pix4D −0.169/0.407 −0.166/1.911 −5.023/12.401 −0.204/1.520 0.061/1.674
Memento Offline −0.105/1.355 −0.071/1.118 −1.135/4.253 −0.389/2.262 −0.114/1.552
above. This shows that if higher accuracy is necessary, a capturing process with-
out a turntable and with uniform lighting and more diverce camera positions
need to be used, even if this will cost more time and resources.
6 Conclusion and Future Work

Our paper presents a head to head comparison of the state of the art SfM 3D
reconstruction software solutions. As part of the research we tested six pro-
grams - ContextCapture, PhotoScan, Memento, Reality Capture, Pix4D and
3DF Zephyr. We tests the programs on both a variety of challenging objects and
on images taken from different capturing conditions. Reconstruction results were
evaluated against ground truth objects on the basis of distance measurement and
roughness comparison.
We demonstrated that programs can be roughly divided in two groups - ones
that are more robust to sub-optimal objects and capturing conditions, but do
not manage to capture smaller details and ones that can capture high amount of
details, but degrade in performance and introduce a lot of noise, once the optimal
conditions are not met. Additionally we show that using a turntable can have
a negative effect on the accuracy of the reconstructed objects, as well as using
a single light source. For optimal capture conditions a moving camera, multiple
lights and images taken from multiple locations and angles are recommended.
As an extension to this paper we propose introducing prior information to
the programs like - camera positions, feature points, markers, etc., as well as
combining multiple software solutions in a pipeline for achieving better results
and helping failed reconstruction attempts on hard to reconstruct objects.
References
1. Bentley: ContextCapture, May 2016. https://www.bentley.com/en/products/
brands/contextcapture
2. Autodesk: ReMake, May 2016. https://remake.autodesk.com/about
3. Agisoft: PhotoScan, May 2016. http://www.agisoft.com/

4. CapturingReality: Reality Capture, May 2016. https://www.capturingreality.com/
5. Pix4D, May 2016. https://pix4d.com/
6. 3DFlow: 3DF Zephyr, May 2016. http://www.3dflow.net/
7. CloudCompare, May 2016. http://www.cloudcompare.org/
8. Izadi, S., Kim, D., Hilliges, O., Molyneaux, D., Newcombe, R., Kohli, P., Fitzgib-
bon, A.: KinectFusion: real-time 3D reconstruction and interaction using a moving
depth camera. In: Proceedings of the 24th Annual ACM Symposium on User Inter-
face Software and Technology, pp. 559–568. ACM (2011)
9. Singh, S.P., Jain, K., Mandla, V.R.: 3D scene reconstruction from video camera
for virtual 3D city modeling. Am. J. Eng. Res. 3(1), 140–148 (2014)
10. Smith, M.W., Carrivick, J.L., Quincey, D.J.: Structure from motion photogram-
metry in physical geography. In: Progress in Physical Geography (2015)
11. Yilmaz, O., Karakus, F. Stereo and Kinect fusion for continuous 3D reconstruction
and visual odometry. In: 2013 International Conference on Electronics, Computer
and Computation (ICECCO), pp. 115–118. IEEE (2013)
12. Schning, J., Heidemann, G.: Taxonomy of 3D sensors. Argos 3, P100 (2016). VISI-
App
13. Ahmadabadian, A.H., Robson, S., Boehm, J., Shortis, M., Wenzel, K., Fritsch, D.:
A comparison of dense matching algorithms for scaled surface reconstruction using
stereo camera rigs. ISPRS J. Photogram. Remote Sens. 78, 157–167 (2013)
14. Kaartinen, H., Hyypp, J., Kukko, A., Jaakkola, A., Hyypp, H.: Benchmarking the
performance of mobile laser scanning systems using a permanent test field. Sensors
12(9), 12814–12835 (2012)
15. Nex, F., Gerke, M., Remondino, F., Przybilla, H.J., Bumker, M., Zurhorst, A.:
ISPRS benchmark for multi-platform photogrammetry. ISPRS Ann. Photogram.
Remote Sens. Spat. Inf. Sci. 2(3), 135 (2015)
16. Koutsoudis, A., Vidmar, B., Ioannakis, G., Arnaoutoglou, F., Pavlidis, G.,
Chamzas, C.: Multi-image 3D reconstruction data evaluation. J. Cult. Heritage
15(1), 73–79 (2014)
17. Schning, J., Heidemann, G.: Evaluation of multi-view 3D reconstruction software.
In: Azzopardi, G., Petkov, N. (eds.) CAIP 2015. LNCS, vol. 9257, pp. 450–461.
Springer, Switzerland (2015). doi:10.1007/978-3-319-23117-4 39
18. Besl, P.J., McKay, N.D.: Method for registration of 3-D shapes. In: Robotics-DL
tentative, International Society for Optics and Photonics, pp. 586–606 (1992)
19. Wang, K., Torkhani, F., Montanvert, A.: A fast roughness-based approach to the
assessment of 3D mesh visual quality. Comput. Graph. 36(7), 808–818 (2012)
20. Guidi, G., Gonizzi, S., Micoli, L.L.: Image pre-processing for optimizing automated
photogrammetry performances. ISPRS Ann. Photogram. Remote Sens. Spat. Inf.
Sci. 2(5), 145 (2014)
21. Nicolae, C., Nocerino, E., Menna, F., Remondino, F.: Photogrammetry applied to
problematic artefacts. Int. Arch. Photogram. Remote Sens. Spat. Inf. Sci. 40(5),
451 (2014)
22. Kullback, S., Leibler, R.A.: On information and sufficiency. Ann. Math. Stat. 22(1),
79–86 (1951)
Proportional Systems in the Design of the Cathedral
of St. George of the Greeks, Cyprus
Douglas Cawthorne ✉ and Romylos Irodotou

( )
The Digital Building Heritage Group, De Montfort University, The Gateway,

Leicester, LE1 9BH, UK
[email protected], [email protected]
Abstract. The cathedral of St. George of the Greeks was built in the 14th–15th c. in
Famagusta, Cyprus to accommodate the religious needs of the Greek orthodox
community living under a Frankish aristocracy. Its design is a hybrid of western
European and Greek orthodox architectural traditions which reflect the political and
social circumstances of its creation. This paper examines the degree to which the
underlying design methods employed can be extrapolated from the physical remains
of the building, the historical sources bearing upon its interpretation and compari‐
sons with related structures. Results are presented of a recent (2016) photogram‐
metric survey of the building and a new digital reconstruction of the church derived
from it. These are used to quantify, assess and illustrate a three dimensional arma‐
ture of regulatory proportions which it is proposed for reasons of ecclesiastical
philosophy and practical execution, were employed to shape the building’s physical
form.
Keywords: Proportions · Medieval · Architecture · Photogrammetry · Geometry ·

Masons · Pell Numbers · Root two · Armature
1 Introduction
The Cathedral of St. George of the Greeks is the largest Orthodox Church from the
Lusignan period (circa 1191–1489) in Cyprus and is of an architectural style that might
best be described as Franco-Byzantine. Now ruinous it was begun probably in the mid
to late fourteenth century as an addition to the much older and smaller Hagios Epiphanios
which adjoins it on the south side. St George’s can be viewed as an orthodox alternative
and some might say a competitor to the Latin cathedral of St. Nicholas, which is situated
a mere 150 m away. St George’s combination of a Latin basilica and sculptural detailing
with Byzantine centralized planning and byzantine religious art raises a number of
questions about its provenance and the motivations and preoccupations of its architects.
In this regard the circumstances of its invention are germane. Until the Crusades in the
twelfth century AD medieval Cyprus was culturally and religiously Byzantine albeit
with a substantial Arab influence. This changed in the 1190’s when the English King
Richard I gave the island as a fiefdom to Guy de Lusignan a long time vassal who had
been ousted from his position as king of Jerusalem after the death of his wife Sybilla [1].
Although described by his peers as “simple and unsophisticated” (“simplex et minus

DOI: 10.1007/978-3-319-48496-9_3
28 D. Cawthorne and R. Irodotou
astutus”) [2] Guy de Lusignan succeeded in holding Cyprus by means of a series of

military and political stratagems (Fig. 1).
Fig. 1. Left - the West end of the Cathedral of St George of the Greeks showing the rose window/
oculus. Right - the East end of the Cathedral showing the three apses, lancet windows and flying
buttresses supporting the clerestory.
In doing so he founded a dynasty of Frankish rulers and fundamentally recast Cypriot

society in a European feudal mold. Guy de Lusignan died in 1194 and was succeeded
by his brother Amaury who obtained a crown from the Holy Roman Emperor, Henry
VI becoming in 1197 the first Lusignan king of Cyprus. Of importance here is that
Amaury took steps to displace the Orthodox Church by introducing the Roman commu‐
nion, a process that was continued by his successors with papal assistance. Cyprus
became the center of European trade with Africa and Asia after the fall of Acre in 1291
and this brought the kings of Cyprus into conflict with Italian merchants who dominated
trade in the eastern Mediterranean. This led to a Genoese occupation of Famagusta in
1372. The kingdom was then made a tributary to the Mameluke state in 1426 and the
succeeding monarchs gradually lost almost all independence until the last Queen, Cath‐
erine Cornaro, was forced to sell the island to Venice in 1472. The Genoese occupation
of Famagusta from 1372 and later the Venetian Republic from 1472 allowed the
Orthodox Church to regain a position it did not have during the feudal kingdom of the
Lusignans, thus despite the absence of evidence regarding its foundation and commis‐
sioning, it has generally been assumed that the building of St George of the Greeks was
started during or shortly after this recrudescence of Eastern Orthodoxy in the mid-four‐
teenth century under the influence of the Genoese [3]. However, as Jeffery noted of this
period in Famagusta’s history in 1918 “not a single monument of importance can be
identified with any certainty as belonging to the period of the Genoese Occupation of
the city” [4, p. 103]. It is within this prolonged condition of uncertainty that we
approached research into the intentions of the architects of the cathedral of St George
of the Greeks.
2 The Surviving Fabric of St George of the Greeks
The Cathedral of St George of the Greeks is in a ruinous, roofless condition and lies in
the south eastern quarter of the old walled city of Famagusta, close to the Latin Cathedral
Proportional Systems in the Design of the Cathedral 29
of St. Nicholas. Unlike St. Nicholas’s which is still in use, now as a mosque, St George’s
retains some of its original internal wall decoration and is largely unmodified, since it
was severely damaged by an earthquake in 1735 and has been a recorded as ruin for at
least the past one hundred and fifty years. Over this period it has received intermittent
academic attention. It was noted by the Marquis de Vogüé and Baron Emmanuel Guil‐
laume Rey in 1860, and was visited by the British architects Edward I’Anson and Sydney
Vacher in 1882–1883 though it was drawings of the West and South elevations of the
Cathedral of St Nicholas in Famagusta that Vacher exhibited at the Royal Academy in
1882. It was not until 1900 that the physical remains of the Cathedral of St George of
the Greeks was first reported to scholarly audiences by the French architectural historian
Camille Enlart who visited Cyprus in 1896 under the auspices of the Ministère de l’In‐
struction Publique et des Beaux-Arts. He drew several medieval structures dating from
the period of the Crusades (including St George of the Greeks) from which were
produced engravings which were published in his L’art gothique et la renaissance en
Chypre [5]. In 1919 George Everett Jeffery published an account of the building noting
that both St Nicholas and St George were “were undoubtedly erected between the years
1373 and 1571” [4] therefore between the start of the Genoese occupation and the end
of Venetian rule, a view which still persists in many quarters. It should be noted that
both Enlart and Jeffery were not able to record the cathedral’s lower details fully, as it
was still covered with debris from the collapsed vaults [6, p. 169]. The basic plan layout
of St. George comprises a nave divided from two side aisles of equal length by two rows
of four columns unequally spaced such that a square is defined by four columns at the
midpoint of the nave. At the eastern end of the nave and each aisle there is a semicircular
apse covered by semi domes that formed the raised sanctuary or bema. Flying buttresses
supported the upper clerestory walls of the main aisle and it is believed a hemispherical
dome on a short cylindrical tower existed above the central crossing, which was observed
and drawn by Cornelis De Bruyn in 1683 [7, pp. 185–186, Fig. 25]. The austere, unar‐
ticulated external wall surfaces at the east end are pierced only by two-centered arch
windows. The western end of the church, with the three doorways still retains a rose
window above the central one. The building is made entirely from high quality ashlar
limestone masonry. Stylistically it has been argued that St George’s presents closer
parallels with churches in the Crusader mainland states than with Frankish or German
counterparts in mainland Europe, which if true may reflect the origins of the citizens its
served [8, pp. 309–310].
3 The Photogrammetric Survey
Several methods exist for the accurate capture of the 3D form of buildings [9]. Unlike
laser scanning which was considered as an alternative means of data capture for this
project [10, pp. 1–25] stereo photogrammetry requires little specialist equipment and so
does not require special customs security clearance or transportation and tends to take
less time. It does require careful planning and execution on the ground but for the
purposes of this study its advantages of simplicity were overriding factors in its selection
as the preferred technique. Stereo photogrammetry uses a series of photographs of the
building taken from different angles and positions that are then processed through speci‐
alized software to produce a detailed 3D surface model and photographic textures over‐
laid upon it. It is not the purpose of the discussion here to rehearse photogrammetric
methodology, this can be found well described comprehensively elsewhere [11–13].
However the capture of textures in particular for this project is of some consequence.
One of the major advantages of stereo photogrammetry over laser scanning is the ability
to generate photorealistic continuous colour textures precisely mapped onto the 3D
point-cloud geometric data (Fig. 2).
Fig. 2. Left – the 3D digital model of the Cathedral as it exists from the North West showing the
photo-textures overlaid upon the surface of the geometric model. Right – the 3D digital model
without photo-textures, viewed from the south east.
This means that images of internal surface decoration like wall paintings are pre-
mapped to their respective points and are as complete as the original photographs. This
potentially allows the religious frescoes and other iconography and their state of pres‐
ervation to be considered simultaneously along with the 3D form of the building, thus
providing a contextually more complete, richer and more integrated data set for exami‐
nation. It also gives the data set a potential use beyond this study firstly for preservation
and conservation of the building and secondly for developing more effective heritage
interpretation material. St. George’s interior surfaces would originally have been
substantially covered with frescoes of ecclesiastical scenes, only some of which still
survive in a faded and damaged state in some of the tombs’ arches and between the
registers of the three apses with pointed semi-domes in the eastern end. Discussion of
the narrative arc of the interior decorative schema is beyond the scope of this study but
photogrammetric recording of its current state carried out for this study may assist in
doing so in the future. For this study the photogrammetric and texture data has been used
to generate orthographic 2D scale drawings of the building which have been the principal
vehicle for examining the dimensional and proportional relationships of the design
(Fig. 3.).
As noted above previous plans of the building were known to have significant incon‐
sistencies so the 3D point cloud model was horizontally sectioned at 1 m above nave
floor level to give an accurate floor plan that conformed exactly to the elevations and
sections. From these a reconstruction was developed based on the photogrammetric
survey, careful on-site measurement and recording of molding profiles and previous
Fig. 3. A longitudinal section of the existing remains of the interior wall of the south aisle with
a reconstruction based on proportional analysis of the arcuation of the nave, clerestory and cupola
superimposed upon it.
interpretation. In this regard we looked to Theophilus A.H. Mogabgab’s extensive study

and restorations of the historic buildings in Famagusta between 1937 and 1939 [14]. He
photographically recorded much of the fabric of St George of the Greeks at this time
but he also made a reconstruction of the windows which we have used here in recon‐
structing the elevations. It is notable that the remains of a surviving clerestory window
on the south nave wall, instead of being smaller as one might expect, in fact has exactly
the same height as those on the side aisles’ walls. In addition, both groups have the same
distance between their sills and the decorative string-course separating the registers. All
have hood moldings which in wetter climates serve to direct rainwater away from the
window opening, but in Cyprus are functionally anachronistic.
4 Dimensions, Geometry and Proportions
The practical or operative craft of masonry has always relied upon geometry for its
effective execution. It was inevitable that this became imbued with meanings and
symbolism, particularly in temples and churches and so by implication the importance
of an internally coherent geometrical schema for religious buildings was often an essen‐
tial part of their perceived spiritual efficacy as well as an aesthetic goal of elegance,
repose and unity in design. The most ancient, and in the fifteenth and sixteen centuries
most revered western authority on architectural design was the first century BC Roman
author Vitruvius who states: “The architect’s greatest care must be that his buildings
should have their design determined by the proportions of a fixed unit. When therefore
account has been taken of the symmetries of the design and the dimensions have been
worked out by calculation, it is then the business of his skill … to produce a proper
balance by adjustment, adding or subtracting from the symmetry of the design …” [15,
p. III. 1. 1.16]. Here symmetry is used not in the modern axial sense but instead to mean
a balance of proportion by use of geometry. Metaphysically this preoccupation with
proportional design became attached to the Christian doctrine of the affinity of souls.
This was a philosophical concept that led firstly to the idea that an inborn sense makes
us aware of harmony, in other words that the perception of harmony through the senses
is possible by virtue of the affinity of our souls. Secondly as articulated by Leon Battista
Alberti (1404–1472) it led to the idea that that a harmonic balance of proportions through
geometry in a building would more effectively convey the communion of ones soul
through prayer to heaven and allow the soul to partake of the vital force which lay behind
all matter and which bound the universe together [16]. This was an old and much vener‐
ated Aristotelian idea and Fra Luca Bartolomeo de Pacioli (c. 1447–1517) in the second
half of the Quattrocento went so far as to say in his book De Divina Proportione (On the
Divine Proportion) published in Venice in 1509 but written between 1496–1498 that
divine functions are of little value if the church (temple) has not been built ‘with correct
proportions’ (‘con debita proportione’). He goes on to say that perfect proportions must
be applied to churches, whether or not the exact relationships are manifest to the
‘outward’ eye [17]. Beyond this the fabric of the well-proportioned late medieval/early
modern church was not only seen as an amplifier of ones prayers it was also a palimpsest
of metaphor and symbolism which was open to those who understood through spiritual
and moral guidance how to interpret and learn from it. For instance at an elementary
level the geometry of the circle representing heaven and the square representing earth
and the geometric union of these two figures in their correct relative vertical positions
in the central crossing and dome of the church was a symbol of heavenly perfection on
earth. From this background the starting point for proportional analysis of this, or indeed
any late medieval or early modern ecclesiastical building is normally detection of the
module, the fundamental unit (mentioned by Vitruvius) used to regulate the dimensions
of the building. Usually this is a divisor of the wall thickness of the nave which itself is
often a multiple of the local unit of measure. At St George of the Greeks the wall thick‐
ness ranges between 1375 mm and 1400 mm. A Lombardic/Venetian foot = 347.73 mm,
therefore four Lombardic/Venetian feet = 1390.92 mm, which is within a 1.1 % error
of the lower range of 1375 mm and 0.5 % of the upper range of 1400 for the main east-
west wall of the nave. The byzantine foot was 312.3 mm (but could be between 308 mm
and 320 mm) which does not give a good multiple fit to the known wall thickness. We
have no data on a Genoese foot but the Genoese palmo = 248 mm again does not yield
plausible whole multiples to fit the wall dimension. However the French ell [aune]
(1372 mm) lies dimensionally very close to the lower limit of the wall thickness. The
French ell was mainly used as a cloth measure and it would be unusual to find it delib‐
erately applied within an architectural context and so for the purposes of this study we
discount it. The columns of the nave originally had the same thickness as the walls
(1400 mm) but were later reinforced with an additional shell of masonry in 1491 after
an earthquake [6, p. 190]. This increased the diameter to ~2300 mm. The triplet
colonettes on the nave wall oppose each column are 680–690 mm (~2 Venetian feet).
Given the above it is recognized here that the closeness of fit of the wall and column
dimensions to Lombardic/Venetian feet may not necessarily indicate a post 1372 date
(i.e. after the Genoese occupation of Famagusta) but could do and could also indicate a
design created by a northern Italian master mason some time before this. In an attempt
to reconcile or “harmonise” the available historical evidence for the building of the new
Cathedral of St George of the Greeks Kaffenberger places its erection between 1350 and
1374 and states that it was “probably begun in around the 1360 s at the very latest” [6,
p. 179] and suggests citizens’ piety following a plague of 1349 as a possible motivating
factor in the Cathedral’s inception. He also notes both that there were substantial dona‐
tions towards the building of the Orthodox cathedral documented in the year 1363 [6,
p. 180] and that, “…in the 1360s, the city was already suffering from a commercial
decline, which would have slowly decreased the availability of financial resources
necessary to start large scale building projects” [6, p. 181]. He also notes that if the
cathedral were finished before the Genoese takeover of Famagusta in 1374 this would
mean a construction duration of 25 years, 10 of which in the (1360’s) may have been in
a period of economic hardship, but during which donations were received. Twenty five
years is a short period of time to complete a building of this size and Kaffenberger
suggests between 50–100 years as realistic for the Latin cathedrals of Famagusta and
Nicosia [18]. One might therefore reasonably assume a similar construction period for
St George of the Greeks, in which case a completion date somewhere in the 1390’s or
1400’s at the earliest might be realistic. This would place its completion squarely within
the Genoese occupation of Famagusta which began in 1372. Paschali gives a date range
of between the late 1360’s and ca. 1400 for the completion of the late Byzantine wall
paintings which covered the two lowest registers of the central apse and the three lowest
registers of the south apse [19, p. 282] (Fig. 4).
Fig. 4. Plans of St George of the Greeks: Left - showing the symmetrical central crossing. Right
– the proportional arrangement of root two (√2) rectangles in process of quadrature which define
the plan and the intercolumniation of the church.
Irrespective of the provenance of the unit of measure the plan conforms to medieval
architectural practice insofar as it appears to be extrapolated from the wall width. The
width across the cathedral between inner wall surfaces is 21.0 m or 15 × the wall module.
The interior length of the nave from the west door to the iconostasis is 22 modules, the
bema has a length of 5 modules from west to east, (excluding the apses). The square at
the central crossing is 7 modules on each side from the column centers, and 6 modules
wide from the inner faces of the original columns (before they were thickened). The
bays in the nave are 5 modules long (west to east) by 7 modules wide between column
centers and those in the aisles are 5 modules long by 4 modules wide from the column
center to the internal surface of the wall (Fig. 5).
Fig. 5. Left - plan of the Cathedral derived from the photogrammetry survey overlaid with the
module grid and Pell Numbers. Right - the quadrature motif, one of a number of representations
of this very specific geometric figure which appear on the internal walls of the church (authors’
emphasis in red lines). (Color figure online)
The use of whole number units in the plan is philosophically attractive in an Aris‐
totelian sense, and would have had a number of practical advantages for the builders
during the cathedral’s construction. However the choice of numbers of units is not arbi‐
trary because they have been selected to give close approximations to the irrational
dimensions that naturally arise from the use of root two, the diagonal of the square. Many
are in fact Pell Numbers which asymptotically approximate root two ratios as their values
tend towards infinity. An example here is the body of the cathedral before the iconostasis
an area which is 15 units wide by 22 units long. A true root two rectangle of the same
width of 15 units would have a length of 21.21 units… 3.7 % shorter. This is small
enough to be undetectable to the human eye but not an insignificant deviation from the
geometric ideal. On the one hand the approximation of irrational dimensions with whole
number ratios in this way has a long history in architectural theory and historical practice
and their use at St George of the Greeks may have been a purely practical one in a
perfectly normal inherited technical tradition which accepted this degree of approxi‐
mation. On the other hand it is clear from the extant fabric that the geometric ideal was
well understood. The plan form has at its heart the square as the generative geometric
figure but the arms of the cross-in-square plan it produces are clearly not squares but are
root two (√2) rectangles, whose long sides are equal to the diagonal of the central square.
This ancient form of geometric rather than arithmetic extrapolation from one figure to
another has a number of well-known properties of self-similarity (it is the basis of the
widely used A paper size system), and traditionally is produced using compass and
squares in a method known as quadrature, involving the rotation of a square about its
center. Intriguingly in the south apse of the sanctuary of St George’s there exist some
of the better-preserved frescoes in the first and second register representing the Depo‐
sition and Entombment on the left side of the window and the Anastasis and Holy
Women at the Empty Tomb on the right. A close inspection of the geometric motif that
runs around the window and at the edges shows a repeating image of the quadrature or
rotating squares. The use of a motif (in this case the quadrature) as a semiotic clue to
the underlying geometry of the design of the fabric of the building as whole is a common
architectural trope intended to assist a viewer who had been suitably instructed to better
appreciate or “read” the building as a message of overall coherence, integration and
unity of design intended by the church architects as a reflection of divine creation.
Symbols and signs were central to Christian thought and so a late medieval Christian
would be alert to their presence and receptive to the messages like this one that they
were intended to convey. But of principal interest here is the coherent way in which a
bilaterally symmetrical Orthodox centralized plan form has been extended through the
consistent use of repeating root two geometry to create a three aisled (one nave and two
aisles) Franco-Byzantine basilica plan form. We suggest here that it indicates that the
intention to unify these two traditions of Christian architectural design was not an after‐
thought or whim but was encoded from the outset in the “genes” of the design when the
architect first started drawing its basic plan form. The mechanics of the process of
extending that initial centralized square to create the basilica plan and maintain the high
degree of proportional unity and coherence in it is only possible because of the unique
properties of self-similarity of the root two rectangle, properties with which medieval
master masons and later architects were entirely familiar both in the Christian west and
the Orthodox east. It is suggested here that it is the use of ad quadratum root two
proportional geometry, which was common to both traditions, that is the real unifying
factor in this building’s design (Fig. 6).
Fig. 6. Left – the internal south wall of the nave taken at the mid line longitudinal section of the
church showing the root (√2) rectangles extrapolated upwards to form a three dimensional
armature. Right – the same section reconstructed based on the root two proportions on the left.
This coherence also extend upwards in the third dimension to the elevational and
sectional arrangement of the design through a relatively straightforward repetition of
identical root two rectangles. The degree of fit between photogrammetry data and arma‐
ture is very good, the height of the registers defined by the decorative string course that
separates the wall into three vertical registers or stories very closely align with multiples
of a five module square giving a good alignment with the top to the springing of the
vaulting. Each of the bays of the vault are of course one × five module squares wide.
The height of the side aisle fits exactly within two × five module squares and the middle
bay within three × √2 rectangular modules. The total height of the nave fits within three
5 × 5 module squares which is equal to the total interior width of the church which is
three by five module squares wide. In order to clarify the overall composition of these
proportional arrangements we recreated them as a three dimensional armature to overlay
upon the 3D data from the photogrammetry as an aid in the digital reconstruction of the
cathedral. It is highly probable that the architects responsible for the design of St George
of Greeks conceived their initial ideas for the building in this way too, as an armature,
a simplified mental model if you like, aided in memory by remembering a few of the
whole number dimensions (lengths, breadth heights and so on) of the armature. From
these key dimensions and the visual picture of their relationships between one another
the architect could extrapolate detail as required, though the property of self similarly
of the root two geometry and the ad quadratum method (Fig. 7).
Fig. 7. The 3D armature of root two (√2) proportions superimposed upon the photogrammetry
data and the reconstruction of the cathedral.
5 Conclusions
The use of root two (√2) geometry and its arithmetical approximation in church archi‐
tecture is by no means unique to St George of the Greeks in Famagusta. Derived from
antecedents in Roman building practice these properties may be found both in western
European church architecture and in churches in the Byzantine world. It was a method
of design shared by medieval and early modern master masons and architects of both
the Orthodox and Latin architectural traditions. It is suggested here that the use of ad
quadratum (the utilization of root two geometry for developing internally coherent
architectural designs) was the key underlying unifying factor in the design of the Cathe‐
dral of St George of the Greeks and that it was recognized as such by the architects. It
shaped the overall form and provided a unifying framework of proportionally related
regulatory lines of control to allow firstly a mathematically and architecturally mean‐
ingful union of orthodox cross-in-square plan with a Latin basilica plan and secondly
an armature on which to place the architectural elements. The method of ad quadratum
was philosophically attractive because through the concept of idealized architectural
proportions it supported the Platonic concept of the affinity of souls and therefore the
efficacy of prayer. There is evidence that a modular system of measurement was used
and that this was applied as arithmetic whole number multiples (Pell Numbers) of the
module to approximate the irrational numbers that naturally arise from the use of root
two geometry. This may have been practically useful during construction but would also
have been philosophically attractive in terms of the Christian interpretation of Aristo‐
telian atomism and the indivisibility of the fundamental unit. There is also evidence from
the surviving wall paintings that this program of design was meant to be understood by
audiences other than the architects, in this case the clerics who served in the sanctuary
beyond the iconostasis. There remains much to discover in the Franco-Byzantine
churches of the Eastern Mediterranean surrounding the role of proportional systems in
religious belief. We would suggest that the approach described briefly here can be
usefully and more fully extended to other buildings of this type to do so.
Acknowledgements. The authors wish to acknowledge the technical and equipment support of
De Montfort University in undertaking this research.
References
1. Flori, J.: Richard the Lionheart: King and Knight. Edinburgh University Press, Edinburgh
(2006). 464 pages
2. Hill, G.F.S.: A History of Cyprus. Cambridge University Press, Cambridge (1972). 554 pages
3. Luttrell, A.: The Latin church in Cyprus, 1195–1312. J. Ecclesiastical Hist. 50 (1999).
Routledge. 380 pages
4. Jeffery, G.E.: A Description of the Historic Monuments of Cyprus. Studies in the Archæology
and Architecture of the Island with Illustrations and Plans, etc., 1918. W. J. Archer, Nicosia
(1919)
5. Enlart, C.: L’art gothique et la renaissance en Chypre: illustré de 34 planches et de 421 figures.
E. Leroux, Paris (1899)
6. Walsh, M.J.K., Kiss, T. Coureas, N. (eds.): The Harbour of all this Sea and Realm: Cusader
to Venetian Famagusta, pp. 171–191. Department of Medieval Studies, Central European
University Press, Budapest, New York (2014)
7. Kaffenberger, T.A., Walsh, M.J.K., Norris, S.J.: Visualising Famagusta: The Virtual
Archaeology of the Orthodox Cathedral of St George of the Greeks, Cyprus (2012)
8. Pringle, D.: Gothic architecture in the Holy Land and Cyprus: from Acre to Famagusta. Levant
47(3), 293–315 (2015)
9. Andrews, D.D., et al.: Measured and Drawn: Techniques and Practice for the Metric Survey
of Historic Buildings, 2nd edn. English Heritage, Swindon (2009). xii, 70 ill. (chiefly col.);
28 cm pages
10. Remondino, F.: Heritage recording and 3D modeling with photogrammetry and 3D scanning.
Remote Sens. 3(6), 1104 (2011)
11. Baltsavias, M., et al. (eds.): Recording, Modeling and Visualization of Cultural Heritage.
Taylor & Francis, London (2006). Edited by E.P. Baltsavias. 528 pages
12. Sapirstein, P.: Accurate measurement with photogrammetry at large sites. J. Archaeol. Sci.
66, 137–145 (2016)
13. Balletti, C., et al.: 3D integrated methodologies for the documentation and the virtual
reconstruction of an archaeological site. In: The International Archives of the
Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XL-5/W4, pp. 215–
222 (2015)
14. Mogabgab, T.A.H., Excavations and improvements in Famagusta. Report of the Department
of Antiquities Cyprus (RDAC), Department of Antiquities Cyprus, pp. 181–90 (1951)
15. Henderson, J., Goold, G.P.: Vitruvius, Vitruvius on Architecture Books I-V. The Loeb
Classical Library, vol. 1. Harvard University Press, London (2002). 322 pages
16. Alberti, L.B., et al.: On the Art of Building in Ten Books, 1st edn. MIT Press, Cambridge
(1988). xxiii, 442 ill.; 27 cm pages
17. Pacioli, L., Da Vinci, L.: De Divina Proportione (On the Divine Proportion): Facsimile (in
Black and White) of the Original Version of 1509. CreateSpace Independent Publishing
Platform (2009). 322 pages
18. Walsh, M.J.K., Edbury, P.W., Coureas, N. (eds.): Medieval and Renaissance Famagusta:
Studies in Architecture. Art and History. Ashgate, Farnham (2012)
19. Paschali, M.: Negotiating identities in fourteenth-century Famagusta: Saint George of the
Greeks, the Liturgy and the Latins, in identity/identities in late medieval Cyprus. Papers given
at the ICS Byzantine Colloquium, pp. 281–301. Centre for Hellenic Studies/Cyprus Research
Centre, King’s College London, Nicosia, London (2014)
The Reconstruction – Argumentation Method
Proposal for a Minimum Standard of Documentation
in the Context of Virtual Reconstructions
Mieke Pfarr-Harfst(&) and Marc Grellert
Digital Design Unit, Technische Universität Darmstadt,

El-Lissitzky-Str. 1, 64287, Darmstadt, Germany
{pfarr,grellert}@dg.tu-darmstadt.de
Abstract. Virtual reconstructions exist for around 25 years. A documentation

of the process of reconstructions was rarely made – a deficit from a scientific
standpoint. One reason was that this was a relatively new discipline and there
was a lack of agreement as to standards and methods. Another was that in many
cases the client did not provide separate funds for a documentation and also did
not require or request them.
In the meantime, many involved parties have become aware of the problem of
the lack of documentation and standards. Besides good scientific practice, also
the guarantee to have access to knowledge embedded in reconstructions should
be realized. However, up to now the proposals orientate themselves rather on
extensive maximal solutions, often coupled with complex data bank applica-
tions, possibly also with annotations to 3D models, which in reality in most
projects would present big challenges as far as usability and available resources
are concerned.
Thus it seemed more constructive to develop a minimal standard, which in
practice would be manageable. The goal of the proposal presented is to compare
images of the reconstruction with the sources and to link them to a written text
(argumentation), which explains upon what basis, including sources, analogies
etc. the reconstruction was made. The core is therefore the triad – “Recon-
struction – Argumentation – Source”. In addition there exists the possibility to
also depict variants for the different areas of a reconstructed building.
The advantage of such a documentation method is that it would be theoret-
ically useable for every kind of architectural reconstruction and thus also for
haptic models, reconstruction drawings or actually built structures. The technical
goal is a web-linked database that can serve as a platform for work, publication
and discussion. The method can also be implemented as a simple text document
with a series of images.
Keywords: Virtual reconstruction Documentation Standards Knowledge

Digital cultural heritage Cultural heritage Visualisation Digital 3D
reconstruction

DOI: 10.1007/978-3-319-48496-9_4
40 M. Pfarr-Harfst and M. Grellert
1 Introduction and Statement of the Problem
At the intersection of architecture, archaeology and architectural history, reconstruc-

tions are a recognised method of visualising and communicating knowledge and the
latest state of research. Classic forms of visualisations include drawings and sketches as
well as haptic models. These have been revolutionised by the rise and steady advance
of the new information and communication technologies. For the last twenty-five years
and more, digital 3D computer model, generally referred to as virtual reconstructions,
have dramatically changed the way reconstructions are done in the cultural heritage
sector. Virtual reconstructions can be seen as direct successors to the traditional
methods of reconstruction, using the new digital media and tools. In the cultural
heritage sector of the humanities, digital processes and applications have become
standard. But virtual reconstructions as a method of visualisation have not only gained
widespread acceptance in the realm of education and knowledge transfer, they have
also become an increasingly widely used research tool in their own right.
The pervasive ubiquity of these digital applications brings with it a new and
unprecedented challenge, namely the long-term storage and protection of this growing
body of digitally stored knowledge and information. The UNESCO [1] spells out the
task in its “Charter on the Preservation of Digital Heritage” [1], warning “that this
digital heritage is at risk of being lost and that its preservation for the benefit of present
and future generations is an urgent issue of worldwide concern” [1].
Similarly, in its “Memorandum zur Langzeitverfügbarkeit digitaler Informationen
in Deutschland” (‘Memorandum on the long-term availability of digital information
Germany’), the Federal Ministry of Education and Research (BMBF) [2] describes the
“preservation of digital objects” as an “undertaking of national significance within the
international context”. It demands that “digital long-term archiving [should encompass]
the entire spectrum of digital objects” and that “transparency of information in the
shape of knowledge mapping should be promoted” [2].
Virtual reconstructions can be seen as both repositories of knowledge and
knowledge models, because they bring together knowledge from heterogeneous
sources and data. The translation of information into 3D calls for a continual con-
cretisation in the interpretation of the underlying facts and assumptions. This inter-
disciplinary process generates new knowledge by means of virtual reconstructions; the
models become carriers and repositories of knowledge. What’s more, they represent the
fusion of different kinds of knowledge that can be described as primary and secondary
sources. Primary sources are findings from excavations; secondary sources comprise
excerpts from literature, topographical surveys, plans, sketches, comparable structures
and, above all, the knowledge accrued by individuals. More often than not, it is this
individual knowledge that is indispensable for the creation of virtual reconstructions.
Based on Mahr [3] it is possible to distinguish three types of knowledge that grow
out of the reconstruction process:
Knowledge inherent in the model: stored knowledge from different sources trans-
lated and transferred into 3D.
Knowledge around the model: knowledge that comprises the context of the model,
important background information about the project, project partners, technical
The Reconstruction – Argumentation Method 41
systems, intention and goals – all of which are factors that have an indirect influence on
the model and the final result.
Knowledge arising from the model: new knowledge that was generated by the
translation into 3D and the bringing together of the different sources [4].
Like all digital data, virtual reconstructions and the multifaceted knowledge they
contain can easily be lost without strategies for their preservation and storage. Even if
the 3D data of the physical model survive, it is safe to assume that the knowledge basis
on which the model had been constructed can no longer be accessed, and that earlier
findings are no longer transparent or comprehensible. To safeguard the sustainability of
the model, it is vital to preserve the technical accessibility of the underlying datasets as
well as the wealth of knowledge on which they were based.
Safeguarding the accessibility of the datasets is a question of sustained backup
strategies and hardware. Moreover, the accessibility and legibility of datasets is
predicated on their unbroken migration from one generation of hard- and software to
the next. This first area of sustainability is not central to this paper, as it is an area that
has to be tackled by computer scientists working closely with the users and creators of
3D models.
This paper, by contrast, addresses the challenges presented by the second area,
namely the preservation of the knowledge that underlies the digital dataset and that
primarily underpins its geometry and texture. The focus here is on the preservation of
the underpinnings in the shape of the heterogeneous sources and on the visualisation of
the process and the decisions that informed it. Drawing on the demands of the London
[5] and Seville Charters [6], appropriate documentation methods and strategies have to
be developed for virtual reconstructions to safeguard the underlying knowledge as well
as the sustainability and verifiability of the models.
More often than not, however, these efforts are jeopardised by insufficient funding
and personnel shortages, which constitute a key reason for the looming loss of
knowledge. Another threat is posed by the lack of uniform documentation systems and
methods. It is here that action is needed. We need to develop everyday methods that
can easily be integrated into the work process, methods that lend themselves to the
annotation of the reconstructed architecture with “footnotes”.
Usability, the simple and effective integration of the method into everyone’s everyday
practice and research, is of paramount importance and has to be integral to the strategy or
system. At the same time, the general accessibility of the documentation and of the tools
developed should be safeguarded, for example through an open access scheme.
It should be emphasised that the development of documentation standards is
indispensable in two areas of virtual reconstructions: in keeping the knowledge up to
date with the latest findings and in upholding academic standards of best practice.
Virtual reconstructions allow for easy updating whenever new data becomes
available, but updating requires the earlier reconstruction process to be traceable and
comprehensible.
If we conceive of virtual reconstructions as a tool of the humanities, the demand for
comprehensive documentation is simply integral to academic standards of best practice.
As in every academic and scientific discipline, good practice requires findings to be
documented to make them comprehensible and available to succeeding generations as
the basis of further research.
2 Status Quo
The exhortation to document the knowledge contained in virtual reconstructions is not

new. The subject has preoccupied the humanities for several years. In light of the
current situation, we can assume that the knowledge that underpins most virtual
reconstructions is not transparent and cannot be accessed or traced. Publications that
document the path of investigation, the findings, the methodology and the sources – i.e.
publications that bear comparison with the documentation of findings in the natural
sciences – are few and far between [7].
In the current research landscape, virtual reconstructions fall within the purview of
the digital humanities, a relatively new area of interdisciplinary enquiry at the inter-
section of computing and the humanities that draws on the steady rise of new infor-
mation and communication technologies in the humanities and the neighbouring
disciplines. Whereas the digital humanities are already well-established in the UK and
the USA, the field is still relatively new in the German academic landscape. The current
state of research can therefore be described as heterogeneous [8].
While the virtual reconstructions as typologies form a subcategory in the field of the
digital humanities, the question of documentation systems or standards is to be
understood as a subtopic of virtual reconstructions.
As early as 2003, the UNESCO Charter on the Preservation of Digital Heritage [1]
called on the research community to develop strategies and solutions to prevent the loss
of knowledge. In the current research landscape, two types of approaches to this
problem can be distinguished. On the one hand, approaches that deal with the problem
on an epistemological level and on the other hand, those that annotate the 3D model
itself with the relevant information and references.
The London [5] and Seville Charters [6] can be read as following the epistemo-
logical approach and classified as theoretical position papers, neither of which presents
concrete solutions. Moreover, because of their limited relevancy to practice thus far
neither charter has taken root and become fully established. Nevertheless, both charters
played an important role in highlighting the problems surrounding virtual reconstruc-
tions and the concomitant lack of rigorous documentation strategies.
There are very few publications that address the subject on an epistemological
basis. A few conference papers touch on it, foregrounding aspects such as hypothesis
and findings and the evidence that underpins them through universally accepted
modelling and notation [9]. What all of these theoretical discussion papers have in
common is that they call for minimum standards of transparency in the documentation
of the processes and sources that underpin virtual reconstructions and for the direct link
of that information with the reconstruction [9, 10].
The doctoral thesis “Dokumentationssystem für Digitale Rekonstruktionen am
Beispiel der Grabanlage Zhaoling” (‘Documentation system for virtual reconstructions
using the example of the Zhaoling burial site’, 2010, Technical University, Darmstadt,
Department of Information and Communication Technology in Architecture) [11] also
belongs to the epistemological line of studies. Starting with the principles of academic
documentation and the special requirements of virtual reconstructions, the author
developed a best practice strategy. The documentation system, which can also be
described as four-level system, covers all important levels for a valid, transparent and
editable documentation. It records not only the context and background of the project,
but also the underlying sources and the reconstruction process. On the evidence level,
the sources are linked to the relevant digital model; the method catalogue presents the
process as an input-output principle. Thus the different states of the model and the
relevant decisions and reasoning that underpins them are made transparent. This the-
oretical system was then tested on a concrete example, giving rise to one of the first
comprehensively documented virtual reconstructions [11, 12].
The research project “Digitales Forum Romanum (‘Digital Forum Romanum’) of
the Winckelmann-Institut of the Humboldt-Universität of Berlin is an example for
making a virtual reconstruction understandable”. Here the virtual reconstruction of
different time periods of Forum Romanum are linked with a WIKI, in which the basis
of the reconstructions is publicly accessible [13].
Alongside the epistemological approach, the current research landscape also fea-
tures projects that investigate the potential of annotating the 3D model with the relevant
information. The spectrum of objectives and themes is wide, but all are driven by the
ambition to find a paradigmatic solution.
The project “Virtual reconstructions in transnational research environments – the
web portal Palaces and Parks in former East Prussia” [14] funded by the Leibnitz
Association uses the 3D model as an interface for navigation, linking it to an under-
lying database by means of semantic annotation and the integration of a wide range of
meta and paradata. This database stores all of the sources and contributors and every
step of the process. Information is entered via complex input masks that draw on an
ontology developed especially for this project [14, 15].
Another, recently completed project in this context is the “DokuVis” [16] docu-
mentation system of the Dresden Technical University. Based on CIDOC CRM [17], a
dedicated cultural heritage ontology, it too is concerned with the annotation of 3D
models with information about sources and the reconstruction process [16].
In this context, we should also mention the “Oplontis project” at King’s College in
London [18], which is based on the same principles, and the reconstruction of the
cloister of Saint-Guilhem-le-Désert [19]. The latter offers an interesting way of doc-
umenting individual architectural elements, their historical context and their position in
a 3D model. For example, it is annotated with information on the current whereabouts
of original architectural fragments of the monastery. A system of colour coding links
the fragment with the museum in which it is now preserved.
It must be noted that most publications tend to treat the subject of documentation in
the context of the project they deal with. They are less concerned with trying to
formulate a proposal for standards of documentation that would serve as a basis for
discussion.
Numerous projects, which aim for the comprehensive annotation of the 3D model,
are based on highly complex systems, e.g. annotated systems [20, 21] or principles of
BIM [22] and ontologies. Although these processes run in the background, the nec-
essary datasets for information about sources, process protocols etc. have to be entered
into the system during the reconstruction process by members of the project team.
Complex input masks with specialised terms and concepts often stand in the way of
quick and efficient data entering and require additional funding and staff. This form of
documentation is only viable in well-funded research projects.
Whether the approach of comprehensive, global standardisation can be shown to
work with ontologies – for example in the context of the Semantic Web – remains, we
believe, debatable. Even advocates of this approach such as Doerr and Iorizzo [23]
stress that ‘solutions’ have to be ‘scalable’. ‘A global model must be relevant, small
and manageable’ [23]. It goes without saying that it is sensible to make the docu-
mentation of the reconstruction process in addition to the reconstruction results and,
possibly, the model itself available to national or European platforms, for example
Europeana, as the EU stipulates [24]. But for the documentation to make sense, it must
be based on a process of consultation. It is only through broad acceptance and prac-
ticable approaches that documentation can be established at all and that, in a second
step, the digital cultural heritage contained in the documentation can be preserved and
made accessible in the long term.
3 New Approach Towards a Minimum Standard

of Documentation - The ‘Reconstruction - Argumentation
Method’
Based on the analysis of the status quo, it appears sensible to develop a minimum
standard that is achievable in everyday practice. The approach presented here aims to
link screenshots or renderings of the virtual reconstruction with the relevant underlying
source information and to tie them together with an explanatory text (argumentation).
The text provides explanations about the bases, sources and analogies used in the
reconstruction. At the heart of this approach is the triad of ‘Reconstruction - Argu-
mentation - Source’ (Figs. 2 and 3). Each area of the reconstructed building is linked
with such a triad. Each of the different areas is tagged and positioned on an overview
images. In addition, there is the option of visualising variants for any number of the
areas. Each ‘Reconstruction - Argumentation - Source’ triad is assigned a value on an
evaluation scale ranging from ‘substantiated’ to ‘probable’, ‘possible’ and ‘hypothet-
ical’. The following diagram illustrates the structure of the documentation method:
Project
• Area 1
Variant 1 Reconstruction - Argumentation - Source
• Area 2
Variant …
• Area 3
• Area …
This type of documentation method has the advantage of being applicable to any kind
of architectural reconstruction, be that a haptic model, a reconstruction drawing or an
actual built structure. It would place the discourse about the scholarly documentation of
reconstruction projects on a broader, more generally accessible foundation. A docu-
mentation method such as this – applicable to all kinds of architectural reconstructions,
image-centred and comparatively undemanding in terms of information technology –
may also help to dispel reservations against virtual reconstructions as opposed to
traditional reconstructions.
The scope of the documentation – in the sense of the number of areas covered – can
be individually defined and adapted to the available resources in terms of time, funding
and personnel.
The technical goal is a web-linked database that can serve as a platform for work,
publication and discussion. In addition to the database application, the method can also
be implemented as a simple text document with a series of images.
Decisions as to which pieces of information should serve as the minimum basis are
facilitated by a combination of data fields and free text input. The latter provides the
opportunity to add any amount of further information. The focus is on textual infor-
mation. The following schematic provides an overview of the information the docu-
mentation should supply.
Fig. 1. Information for each project (example Kaiserpfalz Frankfurt)
Project (Fig. 1):

• Project name
• Runtime of the project
• Institution, under whose umbrella the reconstruction was undertaken
• Person/persons responsible
• Scientific advice
• Contracting body/co-operation partner(s)
• Sponsor
Fig. 2. ‘Reconstruction - Argumentation - Source’ triad for the area ‘Clerestory windows’
(Kaiserpfalz Frankfurt, Church of Saint Saviour)
Fig. 3. ‘Reconstruction - Argumentation - Source’ triad for the area ‘Candelabrum Aula’
(Kaiserpfalz Frankfurt, imperial hall)
• Hardware and software used

• Geographic coordinates of the reconstructed building or urban environment
• Project website
• Name(s) of the contact(s) with email address and phone number
• Further information about the institution: email address, phone number, website,
address and acronym of the institution
• Renderings of the completed project (Fig. 4)
Fig. 4. Project result (example Kaiserpfalz Frankfurt, imperial hall)
Areas
• Number of areas (1: n)
• Name of the area
• Overview image in which the area is mapped, caption
• Number of variants of an individual area (1: n)
Variants
• Name of the variant (standard name ‘1’)
• Assessment of the variants as ‘substantiated’, ‘probable’, ‘possible’ or
‘hypothetical’
• Indication whether or not the variant is part of the final presentation of the project
• Reconstruction – Argumentation – Source
Reconstruction
– Screenshot/rendering (1: n)
– Caption
Argumentation
– Free text input
Source
– Illustration (1: n)
– Caption
– Author
– Date of origin
– Archive
– Signature
– Copyright
– Direct URL if available
– Personal comment for further information
– Type of source (3D Laser scan, SFM architectural survey, drawing of find-
ing, sketch of finding, photographs of the archaeological or architectural
remains, photographs of the existing building, construction plan, proposed
construction plan, survey drawings, haptic reconstruction model, recon-
struction drawing, textual source, historical drawing, historical painting,
historical film)
The method is developed as a database prototype and will be tested at the Department
of Digital Design at the Technical University Darmstadt and by Architectura Virtualis
GmbH, a university’s cooperation partner, in the summer of 2016. The prototype is
available to interested parties who wish to trial it.
References
1. UNESCO: Charter on the Preservation of Digital Heritage, Records of the General
Conference, Paris, 29 September–17 October 2003
2. BMBF und Nestor (eds.): Memorandum zur Langzeitverfügbarkeit digitaler Informationen in
Deutschland (2003). http://www.zlb.de/aktivitaeten/bd_neu/heftinhalte2006/DigitaleBib010506.
pdf
3. Mahr, B.: Das Wissen im Modell. http://www.tu-berlin.de/fileadmin/fg53/KIT-Reports/r150.
pdf
4. Pfarr-Harfst, M.: Typical workflows, documentation approaches and principles of 3D digital
reconstruction of cultural heritage. In: Münster, S., Pfarr-Harfst, M., Kuroczynski, P.,
Ioannides, M. (eds.) 3D Research Challenges in Cultural Heritage. How to Manage Data and
Knowledge Related to Interpretative Digital 3D Reconstructions of Cultural Heritage, vol. 2,
Heidelberg (forthcoming)
5. Denard, H.: The London Charter for the computer-based Visualisation of cultural heritage.
Draft 2.1, 7 February 2009. http://www.londoncharter.org/fileadmin/templates/main/docs/
london_charter_2_1_en.pdf
6. International Forum of Virtual Archaeology (eds.): Principles of Seville. International
Principles of Virtual Archaeology, Sevilla (2014). http://www.arqueologiavirtual.com/
carta/?page_id=437
7. Münster, S.: Entstehungs- und Verwendungskontexte von 3D-CAD-Modellen in den
Geschichtswissenschaften. In: Meissner, K., Engelien, M. (eds.) Virtual Enterprises,
Communities and Social Networks, pp. 99–108. TUDpress, Dresden (2011)
8. Sander, M. (eds.): Aktuelle Herausforderungen im Kontext digitaler Rekonstruktion. Beitrag
der Arbeitsgruppe Digitale Rekonstruktion des Digital Humanities im deutschsprachigen
Raum e.V. zum Agendaprozess, “Zukunft sichern und gestalten” des BMBF. Dresden
(2014). http://digitale-rekonstruktion.info/wp-content/uploads/2015/01/140831_Herausfor
derungen_Digitaler_Rekonstruktion.pdf
9. Grellert, M., Haas, F.: Sharpness versus uncertainty in ‘complete models’. Virtual
reconstructions of the Dresden castle in 1678. In: Hoppe, S., Breitling, S. (eds.) Virtual
Palaces, Part II Lost Palaces and their Afterlife, Virtual Reconstruction between Science and
Media, Munich, pp. 119–148 (2016). http://www.courtresidences.eu/index.php/publications/
e-Publications/#Volume%203
10. Bentkowska-Kafel, A., Denard, H., Baker, D. (eds.): Paradata and Transparency in Virtual
Heritage. Ashgate, Farnham (2012)
11. Pfarr, M.: Dokumentationssystem für digitale Rekonstruktionen am Beispiel der Grabanlage
Zhaoling, Provinz Shaanxi, China. TUprints, Darmstadt (2010). http://tuprints.ulb.tu-
darmstadt.de/2302/
12. Pfarr-Harfst, M.: A new documentation system in the context of cultural heritage. In:
Verhagen, P.U.A. (eds.) Across Space and Time. Proceedings of the 41th International
Conference on Computer Applications and Quantitative Methods in Archaeology (CAA),
Perth, Australia
13. Digitales Forum Romanum. http://www.digitales-forum-romanum.de/
14. Virtual reconstructions in transnational research environments - The portal: palaces and
parks in the former East Prussia. http://www.herder-institut.de/index.php?id=585
15. Kuroczyński, P., Hauck; O., Dworak, D.: Digital reconstruction of cultural heritage –
questions of documentation and visualisation standards for 3D content. In: 5th International
Euro-Mediterranean International Conference on Cultural Heritage (EuroMed 2014),
Cyprus, Heidelberg (2014). http://www.academia.edu/9189049/
16. Bruschke, J., Wacker, M.: Neuartige Werkzeuge für die Entwicklung und Dokumentation
digitaler Rekonstruktionen. In: TU Dresden Forschungsförderung und Transfer, Technolo-
gieZentrumDresden GmbH, Industrie- und Handelskammer Dresden, GWT-TUD GmbH
(eds.) Dresdner Transferbrief, p. 9. TUDpress, Dresden (2015)
17. The CIDOC Conceptual Reference Model. http://www.cidoc-crm.org/
18. The Oplontis Visualization Project. http://www.oplontisproject.org/index.php/people/the-
oplontis-visualization-project/
19. De Luca, L.: A complete methodology for the virtual assembling of dismounted historic
buildings. Int. J. Interact. Des. Manuf. (IJIDeM) (2014). Heidelberg
20. MayaArch3D. http://www.mayaarch3d.org/language/en/sample-page/
21. VSim. https://idre.ucla.edu/research/active-research/vsim
22. Fai, S., Graham, K., Duckworth, T., Wood, N., Attar, R.: Building Information Modelling
and heritage Documentation (2011). https://pdfs.semanticscholar.org/b76a/6be1ab4c5c14
8757eac8c73dc7a5c2995999.pdf
23. Doerr, M., Iorizzo, D.: The dream of a global knowledge network - a new approach.
Comput. Cult. Heritage 1(1), 1–23 (2008)
24. European Commission (eds.): European Commission’s Report on Digitisation, Online
Accessibility and Digital Preservation of Cultural Material (2014). https://ec.europa.eu/
digital-single-market/en/news/european-commissions-report-digitisation-online-accessibility-
and-digital-preservation-cultural
Multi-scale 3D Modelling of Damaged Cultural Sites:
Use Cases and Image-Based Workflows
Styliani Verykokou ✉ , Anastasios Doulamis, George Athanasiou,

( )
Charalabos Ioannidis, and Angelos Amditis

Institute of Communication and Computer Systems, Athens, Greece
[email protected], [email protected],
[email protected],
{george.athanasiou,a.amditis}@iccs.gr
Abstract. The creation of 3D models of Cultural Heritage (CH) sites that have
undergone a severe disaster due to a catastrophic incident (e.g., earthquake,
explosion, terrorist attack) is of great importance for several use cases. Different
actors, like Urban Search and Rescue crews, structural, civil and surveying engi‐
neers, people in charge of restoration plans, archaeologists, architects, reporters,
television presenters and computer engineers, may exploit the 3D information in
a different way. Hence, each of them needs models of different scales/levels of
detail and under different time constraints. In this paper the need for multi-scale
3D models of severely damaged or collapsed CH sites is addressed and various
use cases are discussed. Also, image-based workflows are established for creating
multi-scale 3D products via UAV images of a damaged church due to an earth‐
quake. The models of different scales require very different amounts of time for
their generation and may be used for search and rescue, damage assessment,
geometric documentation, planning of repair works and simple visualization.
Keywords: 3D model · Multi-scale model · Image-based modelling · Cultural

heritage · Disaster · Search and rescue · Photogrammetry
1 Introduction
Up to now, considerable research has been conducted on 3D modelling of Cultural

Heritage (CH) sites using various methods. Terrestrial laser scanning in combination
with photogrammetric and surveying techniques have been extensively used for the
creation of detailed 3D models of historic monuments, heritage works of art and cultural
sites [1–10], for their geometric recording and documentation. Also, thanks to recent
developments in photogrammetry and computer vision, which permit the automated
processing of image datasets, several papers have dealt with image-based 3D modelling
of CH sites without the use of laser scanning technologies, via terrestrial images and/or
imagery from Unmanned Aerial Vehicles (UAVs) [11–13], as well as using large web-
based collections of images [14, 15]. Moreover, some research has been conducted
towards the detection and evaluation of structural damages and surface weathering
damages in CH sites via 3D modelling [16–20]. However, state-of-the-art automated

DOI: 10.1007/978-3-319-48496-9_5
Multi-scale 3D Modelling of Damaged Cultural Sites 51
techniques have not been applied for 3D modelling of severely damaged or collapsed
CH sites due to a natural or man-made disaster (e.g., earthquake, explosion, terrorist
attack); such incidents may cause large-scale damages in CH sites and their 3D recording
is of great importance for several applications.
The most important use case of 3D models of CH sites that have undergone a
destructive incident is the assistance of the work of first responders and Urban Search
and Rescue (USaR) crews in an emergency scenario, in order to detect and locate
possible entrapped victims. It is true that CH sites are highly visited places that attract
a great number of tourists; thus, the availability of a rescue plan in case of an emergency
is particularly important for such sites. The 3D model of a disaster scene may aid the
localization of trapped humans, according to the operations framework proposed by the
EU FP7 project INACHUS [21]. Except for USaR support, the 3D models of damaged
CH sites may be used for damage assessment, geometric documentation, planning of
the restoration processes and simple visualization. In this paper the need for 3D models
of severely damaged CH sites is highlighted and various use cases are discussed.
Furthermore, different actors, like USaR crews, engineers and archaeologists, who
exploit the 3D information of disaster CH scenes in a different way, need 3D models of
different scales/levels of detail, under different time constraints. Image-based modelling
workflows are established for the creation of products at various scales, adapted to the
needs of different professionals. Whereas 3D models can be created using other tech‐
niques as well, like surveying methods and laser scanning technologies, in combination
with image-based processes, emphasis is given on the proper parameterization of the
image-based workflow that leads to the generation of multi-scale 3D models.
2 Use Cases of 3D Models of Damaged Cultural Sites
In this section, the use cases of 3D models of CH sites affected by a disaster are presented,
giving emphasis on the most important one, which lies in the field of USaR.
2.1 Urban Search and Rescue
The most important and immediate need for 3D modelling of CH sites affected by natural
hazards (e.g., a severe earthquake) or man-made disasters with lots of damages and
injured people lies in the fields of disaster response and USaR. CH places (e.g., old
churches, ancient odeons, archaeological sites, historical buildings) are centers of attrac‐
tion for tourists, usually being visited by a great number of people every day. Thus, in
case of a catastrophic incident, there is a high possibility of trapped people. In such an
emergency situation, USaR crews have to make quick decisions under stress to deter‐
mine the location of any trapped alive humans, with the intent of saving lives. The fast
creation of the 3D model of a damaged CH site would significantly assist the work of
USaR teams, increasing their effectiveness in locating people under rubble piles.
This innovative idea is the basic concept of the EU FP7 project INACHUS, which
aims at achieving significant time reduction in the USaR phase by establishing a system
that provides wide-area situation awareness solutions for improved detection and
52 S. Verykokou et al.
localization of trapped victims, assisted by simulation tools for predicting structural

failures and a holistic decision support mechanism incorporating operational procedures
and resources of relevant actors. A fundamental part of the INACHUS framework is the
3D modelling of partially or fully collapsed buildings and, generally, disaster scenes,
using UAV images in combination with terrestrial and airborne laser scanning techni‐
ques. Whereas a live video from a UAV camera could also aid the work of USaR crews,
the 3D models generated in near real-time additionally enable comparison of the actual
disaster scene with simulation results (i.e., a library of collapsed building models) in
order to find possible survival spaces and rescue paths, thereby facilitating the prioriti‐
zation and planning of USaR activities. These outputs may be used by human operators
for visual inspection in order to identify main disaster areas and make critical decisions
with the aim of rapidly locating and saving trapped victims. In parallel with this macro‐
scopic surveillance of the crisis area via simulation tools, imaging techniques and
photogrammetric workflows, a further decrease in the overall rescue time of USaR
operations in a microscopic/per-building level can be achieved by using sensors that are
able to penetrate deeper into the rubble cavities, like autonomous snake robots with
thermal cameras that provide preliminary indication of presence of trapped victims.
Other methods for prioritization of search areas by means of triangulating the location
of radiation originating from personal handsets (mobile phones) buried under rubble can
also be implemented. Hence, the smart combination of 3D modelling pipelines, building
damage simulations, sophisticated data fusion techniques and intelligent inference
methods may help to quickly gain an overview of the actual damage and identify possible
survival spaces. Also, in combination with complementary technologies, USaR forces
can make well-educated decisions and efficiently perform their tasks. More information
on the INACHUS approach can be found in [22].
This idea may be of great significance in CH sites which attract many people and
may collapse partially or fully due to a devastating incident. In such a situation, the near
real-time 3D modelling of a destroyed heritage site may enhance the operational effec‐
tiveness of all those involved in the complex USaR and first response activities, in
combination with other techniques, as part of the INACHUS operations framework.
2.2 Post-disaster Damage Assessment

Another use case of 3D models of damaged CH sites is structural damage assessment,
which is always a priority after the detection and rescue of any trapped victims, espe‐
cially in CH sites of high significance. It has been shown that geometric parameters
derived from 3D point clouds of damaged structures can reveal damage indicators that
are difficult to be detected via the original images, assisting the identification of severely
damaged or completely destroyed buildings [23]. The identification of damage in build‐
ings based on gaps in point clouds has also been addressed for damage assessment [24].
This research is conducted as part of the EU FP7 project RECONASS [25], aiming at
the assessment of the structural condition of facilities after a disaster. The structural
seismic damage assessment is conducted via multi-perspective oblique airborne images
[26]; such kind of images is preferred in comparison with vertical ones for damage
assessment, as they provide visual information for both horizontal and vertical structures
[27]; thus, current research focuses on their metric exploitation, both in terms of exterior
orientation estimation [28] and Dense Image Matching [29]. In this context, 3D model‐
ling via UAV oblique images, could also be applied for structural damage analysis of
damaged CH sites, following various types of catastrophic events.
2.3 Geometric Documentation

The detailed 3D geometric documentation of damaged CH sites may be used as a back‐
ground by engineers and other scientists in order to find the reasons to what caused the
destruction of structures and draw general conclusions. The 3D geometric documenta‐
tion is indispensable for the full registration of the structure deformations [30] and has
to be conducted after the crucial near real-time support of rescue forces and the subse‐
quent damage assessment, also aiding the planning of restoration works (Sect. 2.4).
2.4 Planning of the Restoration Process
The precise 3D modelling of damaged CH sites is essential for planning repair works
and may assist the decision making process concerning the restoration actions to be
undertaken. The availability of a 3D model of a CH site depicting its actual situation
after a severe disaster would assist architects and engineers in charge of restoration and
rehabilitation works. The planning of the restoration process aided by the 3D modelling
products is a significant task for the preservation of the built CH, preventing its loss.
2.5 Photorealistic Visualization
The textured 3D mesh model of a damaged CH site may be used for photorealistic
visualization by newscasts and information programs in television. The fast access of
the viewers to such a model could inform them in a realistic manner about the magnitude
and the extent of the disaster. A photorealistic 3D model of a damaged CH site may also
be used in an augmented reality application [31] for the interactive visualization of
different states of the site throughout the years, e.g., before the disaster, just after the
disaster and the present situation, which ideally corresponds to the reconstructed site.
Similarly, such a model may be used in a virtual reality application or for simple visu‐
alization. These applications are of less importance compared to USaR assistance,
damage assessment, geometric documentation and planning of reconstruction works;
however, they cannot be ignored in today’s multimedia and visualization era.
3 Multi-scale 3D Modelling
Taking into account the different use cases of 3D models of damaged CH sites, it is
evident that several actors of various specialties require such products, including first
responders in an emergency scenario, USaR teams, structural, civil and surveying engi‐
neers, archaeologists, architects, reporters, television presenters, computer engineers,
graphic designers, etc. Each one has diverse requirements and information needs
concerning the resolution and the accuracy of the 3D model as well as the timescale.
Thus, the scale dimension needs to be handled as well in the modelling of damaged CH
sites. The concept of multi-dimensional modelling has already been introduced in liter‐
ature, mainly concerning Land Information Systems [32, 33]; also, its use has been
proposed for large-scale CH resources [34]. In this paper, the scale dimension is
proposed to be integrated in the modelling of damaged CH sites. Models of different
scales correspond to different accuracy, point cloud density and number of triangles or
polygons and may have been produced via different methodologies; e.g., combination
of surveying, laser scanning and photogrammetric techniques for the derivation of a high
accuracy 3D model for geometric documentation or image-based techniques for the
creation of a non-georeferenced 3D model of arbitrary scale, for visualization purposes.
3.1 Different Levels of Detail

Whereas the establishment of a categorization of levels of details (LoDs) for CH sites
requires the cooperation of scientists of various specialties, being out of the scope of
this paper, an indicative differentiation of LoDs for CH sites of the same type as the case
study (i.e., churches, historical buildings, etc.) is proposed, using five LoDs. LoD0 is
proposed to represent the volume of the CH building or its rubble if it has collapsed,
without texture or differentiated surfaces (e.g., openings, columns). LoD1 may represent
a rough 3D textured model, encompassing roof structures, openings and the basic shape
of the rubble. LoD2 may correspond to a low-scale textured 3D model representing the
walls, columns, spaces and other basic structures of the monument. LoD3 is proposed
to represent a medium-scale textured model comprising 3D information about compo‐
nent elements of columns, doors, floor, roof and other architectural elements, also
allowing more precise identification of the rubble and damaged items. Finally, LoD4
may represent a large-scale textured 3D model of a CH site with all the architectural
details, depicting even the smallest cracks and damages caused by a catastrophic inci‐
dent. The models of different LoDs may be organized in a database, possibly in a multi-
level pyramid framework, each one corresponding to a different user’s category.
LoD0 models of damaged CH sites are not useful for the aforementioned applica‐
tions, as they can only serve for a very rough estimation of the volume and location of
the rubble. LoD1 and LoD2 models may be useful for a refined estimation of the volume
and position of the rubble as well as for USaR tasks, because of the short time required
for their generation. LoD2 and LoD3 models may be used for visualization purposes,
as they have a relatively small size. LoD4 models are needed for geometric documen‐
tation, damage assessment and planning of repair works.
3.2 Image-Based Modelling Workflow
Multi-scale models may be produced via different parameterization of the automated

image-based 3D modelling workflow. Its first step is the Structure from Motion (SfM)
method, which refers to the process of simultaneously recovering the camera poses and
3D structure from a 2D image sequence. Feature extraction and matching algorithms
are applied, usually combined with a sequential algorithm for metric reconstruction, in
case of known interior orientation, or projective reconstruction, in case of uncalibrated

cameras; in the latter case, an auto-calibration process is applied. The final step of the
SfM process is usually a bundle adjustment. The georeferencing of the SfM sparse point
cloud and the camera projection centers is generally performed by estimating the 3D
similarity transformation between the arbitrary SfM coordinate system and the world
reference system via control points and/or GPS measurements. The generation of a dense
point cloud using the SfM results is accomplished through Dense Image Matching
(DIM), via a stereo or a multi-view approach. Finally, the dense point cloud may be
converted into a 3D mesh model, which can be textured via the oriented images.
4 Case Study: The Church of Saint Paul in Mirabello
In this section, an example of 3D modelling at various scales via photogrammetric tech‐

niques is presented. The case study is the Church of Saint Paul in Mirabello, Italy, built
in 1795–1804, demolished in 1929, reconstructed in 1943 and partially collapsed due
to the earthquake of May 20, 2012 that hit the Emilia-Romagna region of Italy. Emphasis
is given on the image-based automated 3D modelling workflow and the parameterization
that leads to the creation of models of various LoDs. Surveying methods, possibly laser
scanning techniques and a thorough editing of the resulting point cloud and mesh are
also required for creating accurate models of non-arbitrary scale, combined with photo‐
grammetric methods. However, such techniques are out of the scope of this paper, which
examines the possibility of automatic image-based creation of 3D models at various
LoDs via different parameters, using the same dataset. The modelling pipeline using
UAV images of the Church and the parameters used for the creation of models of
different resolutions are established using one of the most commonly used software.
4.1 Data Capturing and Processing
Dataset Description. Α dataset of images (Fig. 1) depicting the Church of Saint Paul
in Mirabello, acquired using a VTOL (vertical take-off and landing) UAV after the
earthquake of 2012 in Northern Italy, was used. The dataset consists of 131 low-height
images from several perspectives (subset 1) and 79 mainly vertical aerial images (subset
2), captured by an Olympus E-P2 digital camera featuring a 12.3-megapixel sensor
(4032 × 3024 pixels), corresponding to a focal length of 17 mm. The fact that the flight
planning and the planning of image capturing are performed quickly and under stress in
such situations may result in incomplete datasets. The specific dataset lacks of complete‐
ness, as it does not include any oblique images to provide a connection between the low-
height (almost horizontal) and vertical images. Also, the images do not have GPS/INS
information; hence, the generated models are not georeferenced. However, the availa‐
bility of such datasets up to now is quite rare and the specific one provides a represen‐
tative instance of multi-scale image-based 3D modelling of disaster CH sites.
Used Software and Computer Equipment. The Agisoft PhotoScan software [35]
(professional edition, version 1.2.3) was used for the 3D modelling of the damaged
Fig. 1. Images from subsets 1 (top) and 2 (bottom)
Church of Saint Paul; it is a stand-alone commercial software that generates 3D models

from images. Recent research that focused on the evaluation of commercial and open-
source software for 3D modelling of disaster scenes [36] has shown that PhotoScan leads
to satisfactory results in a small time period. It relies on SfM and DIM techniques to
generate a dense point cloud of the scene and applies meshing and texture mapping
algorithms for the creation of complete photorealistic 3D surface models. In addition,
it includes limited point cloud and mesh editing capabilities. The tests were performed
on a 64-bit Intel Core i7-4700MQ 2.40 GHz laptop computer with 24 GB of RAM.
4.2 Image-Based 3D Modelling at Different Scales

In this section, the workflow and the parameterization applied for image-based model‐
ling at various scales via the PhotoScan software are presented and the corresponding
use cases, depending on the resolution of the 3D model, are outlined.
Workflow and Parameter Tuning. 3D point clouds of the two image subsets are
generated independently from each other, because not enough tie points are detected
between them, making it impossible to create a 3D model of the whole Church in one
go. This problem could be solved if oblique imagery were also captured. The dense
clouds are aligned and merged for the generation of a complete model of the Church.
Alignment of Images. The first step is the alignment of images; it is a SfM process that
generates a sparse point cloud of the scene and computes the camera interior and exterior
orientation of every image. A modification of the SIFT algorithm [37] is used for the
extraction of interest points. The alignment of the images of both subsets is performed
using the lowest (ALL), low (AL) and medium (AM) accuracy options available in
PhotoScan for the generation of LoD1, LoD2 and LoD3 3D models respectively, as well
as the high (AH) and highest (AHH) settings for the creation of LoD4 models. ALL
causes image downscaling by a factor of 64 (8 times by each side); at AL the images
are downscaled by a factor of 16; at AM they are downscaled by a factor of 4; at AH
the images of original dimensions are used; at AHH the images are upscaled by a factor
of 4. The latter setting is used for more accurate localization of tie points. Higher accu‐
racy settings are used for more accurate orientation estimation, whereas lower accuracy
options are used to obtain rough camera parameters in a shorter period of time. In order
to speed up the alignment process, the “generic” pair preselection mode is chosen,
according to which the overlapping image pairs are selected via an initial matching
process with a lower accuracy setting. The maximum number of feature points and tie
points on every image is set to 40000 and 1000 respectively.
Dense Point Cloud Generation. The generation of a dense point cloud using the esti‐
mated camera poses is the second step. PhotoScan calculates depth information for each
camera and combines it into a single dense cloud. The dense point cloud creation using
both subsets of images is performed via the lowest (DLL) quality setting for the gener‐
ation of LoD1 and LoD2 3D models, the low (DL) quality setting for the generation of
a LoD3 model as well as the medium (DM) and high (DH) options for creating LoD4
models. DLL implies image size downscaling by a factor of 256 (16 times by each side);
DL downscales images by a factor of 64; at DM images are downscaled by a factor of
16; DH downscales images by a factor of 4.
Merging of “Chunks”. Six common points are manually measured in five images of
each subset for the alignment (A) of the “chunks” of images and the dense point clouds
via the estimation of the 3D similarity transformation between them. Some wrong and
noisy cloud points are manually deleted, to obtain a cleaner mesh in the subsequent stage
of meshing. The aligned “chunks” are merged, resulting into 210 oriented images in a
common coordinate system; according to this process (M), the point clouds are
combined into a single dense cloud.
Meshing. The multi-scale dense point clouds are transformed into meshes. The “arbi‐
trary” surface type is chosen for all models and the maximum number of polygons is set
to 1/5 of the number of dense cloud points, using the “high” setting (MH). The inter‐
polation mode is enabled to automatically fill some holes in areas without points.
Texture Mapping. Texture mapping is the final step. The “generic” mapping mode,
which does not make any assumptions on the scene type, and the “mosaic” mode, which
performs blending of the low frequency component for overlapping images and uses the
high frequency component from one image, are applied. Color correction is disabled.
The texture size is set to 5000 × 5000 pixels for the LoD1, LoD2 and LoD3 models (T5)
and 15000 × 15000 for the LoD4 models (T15). According to the aforementioned steps,
the multi-scale model of the damaged Church is created. Specifically, five 3D models
of different scale that correspond to LoD1 – LoD4 are generated using the same dataset;
the parameterization used is summarized in Table 1.
Table 1. Parameterization used for the generation of 3D models at different LoDs

LoD Parameterization
LoD1 ALL + DLL + A + M + MH + T5
LoD2 AL + DLL + A + M + MH + T5
LoD3 AM + DL + A + M + MH + T5
LoD4 AH + DM + A + M + MH + T15 / AHH + DH + A + M + MH + T15
Results and Discussion. The computational time of the first two steps (i.e., alignment
of images and DIM), which are the most time consuming ones, and the number of dense
cloud points for each subset of images are presented in the form of clustered bar charts
in logarithmic scale in Fig. 2 (top). Figure 2 (bottom) indicates the total computational
time of 3D modelling for the combined image subsets (entire dataset) in logarithmic
scale. Both the computational time of SfM and DIM as well as the total 3D modelling
time and the number of generated points increase approximately exponentially in each
higher parameterization setting. The effectiveness in terms of time is very crucial for
emergency situations, like the one addressed by the INACHUS project. In such situa‐
tions, a coarse model of the scene has to be created in a very short time period in order
to be used by USaR teams to locate survivors, in combination with other techniques.
Thus, the 3D model generated via the ALL + DLL + A + M + MH + T5 settings in
near real-time can be used as a first rough LoD1 model; subsequently, the AL + DLL + A
+ M + MH + T5 parameters may be used for the creation of a low-scale LoD2 model
in a short time period. The creation of a LoD3 model for such situations is advised if a
powerful computer is used, only after the faster generation of a LoD2 model. LoD2 and
LoD3 models may also be used for visualization purposes, e.g., for newscasts and
augmented reality applications; especially, the latter types of applications need models
of small size, i.e., low resolution, because of their requirement for real-time rendering.
LoD4 models (e.g., via the AH + DM + A + M + MH + T15 settings) may be used for
damage assessment and LoD4 models of the highest possible resolution (e.g., via the
AHH + DH + A + M + MH + T15 settings) may be used for geometric documentation
of damaged CH sites and planning of restoration works. In any case, much attention has
to be paid in the image acquisition process, which has to be held by a professional, so
that complete datasets are obtained, avoiding data redundancy; the latter is very crucial
for near real-time applications (e.g., USaR). Figure 3 shows details of the dense point
Fig. 2. Bar charts indicating the computational time of the SfM and DIM steps for each subset
of images (top, left), the number of generated cloud points for each subset (top, right) and the
computational time required for the generation of a textured 3D model (bottom) via different
settings
clouds used for the creation of the multi-scale model of the Church of Saint Paul.
Figure 4 illustrates views of the non-textured and textured meshes of the Church at
different LoDs.
Fig. 3. (a)–(e): Views of the 3D dense point clouds of increasing scale (ALL + DLL, AL + DLL,
AM + DL, AH + DM, AHH + DH) used for the creation of the multi-scale 3D model of the
damaged Church of Saint Paul; (f)–(g): details of the AH + DM (f) and AHH + DH (g) dense
point clouds
Fig. 4. 3D models of the damaged Church of Saint Paul at different scales; from left to right:
LoD2 model (AL + DLL + A + M + MH), LoD3 model (AM + DL + A + M + MH), LoD4
model (AH + DM + A + M + MH), LoD4 textured model (AH + DM + A + M + MH + T15)
5 Conclusions
In this paper the generation of multi-scale 3D models of CH sites seriously damaged

due to a catastrophic incident has been addressed. The most important use case of such
models lies in the field of search and rescue during an emergency scenario. CH sites are
highly visited by many people, so the detection of possible victims has to be held using
a well-defined plan; the 3D model of the disaster scene may contribute significantly to
the localization of entrapped humans. Other use cases lie in the fields of damage assess‐
ment, geometric documentation, planning of repair operations and photorealistic visu‐
alization. Image-based 3D modelling workflows have been established for the afore‐
mentioned use cases, resulting in models of different scales, which require very different
amounts of computational time for their generation. All the 3D models of different scales
of the same CH site form a multi-scale model of the site, which encompasses various
levels of detail, each one corresponding to a different application.
Acknowledgements. This work was supported by the European Commission under INACHUS,
a collaborative project, part of the FP7 for research, technological development and demonstration
(grant agreement no. 607522).
References
1. Ioannidis, C., Tsakiri, M.: Laser scanning and photogrammetry for the documentation of a
large statue - experiences in the combined use. In: CIPA XIX International Symposium, pp.
517–523 (2003)
2. Sofocleous, E., Ioannides, M., Ioannidis, C.: The geometric documentation of the Asinou
Church in Cyprus. In: VAST 2006, pp. 138–144 (2006)
3. Guarnieri, A., Remondino, F., Vettore, A.: Digital photogrammetry and TLS data fusion
applied to cultural heritage 3D modeling. ISPRS Arch. 36(Part 5) (2006). 6 pages
4. Lambers, K., Eisenbeiss, H., Sauerbier, M., Kupferschmidt, D., Gaisecker, T., Sotoodeh, S.,
Hanusch, T.: Combining photogrammetry and laser scanning for the recording and modelling
of the late intermediate period site of Pinchango Alto, Palpa. Peru. J. Archaeol. Sci. 34(10),
1702–1712 (2007)
5. Remondino, F., El-Hakim, S., Baltsavias, E., Picard, M., Grammatikopoulos, L.: Image-based
3D modeling of the Erechteion, Acropolis of Athens. ISPRS Archives 37(Part B5), 1083–
1092 (2008)
6. Rüther, H., Chazan, M., Schroeder, R., Neeser, R., Held, C., Walker, S.J., Matmon, A.,
Horwitz, L.K.: Laser scanning for conservation and research of African cultural heritage sites:
the case study of Wonderwerk Cave. South Africa. J. Archaeol. Sci. 36(9), 1847–1856 (2009)
7. Fregonese, L., Taffurelli, L.: 3D model for the documentation of cultural heritage: the wooden
domes of St. Mark’s Basilica in Venice. ISPRS Arch. 38(Part 5) (2009). 6 pages
8. Lerma, J.L., Navarro, S., Cabrelles, M., Villaverde, V.: Terrestrial laser scanning and close
range photogrammetry for 3D archaeological documentation: the upper Palaeolithic Cave of
Parpalló as a case study. J. Archaeol. Sci. 37(3), 499–507 (2010)
9. Ioannidis, C., Valanis, A., Tapinaki, S., Georgopoulos, A.: Archaeological documentation and
restoration using contemporary methods. In: CAA2010, pp. 1–4 (2010)
10. Lerma, J. L., Seguí, A. E., Cabrelles, M., Haddad, N., Navarro, S., Akasheh, T.: Integration
of laser scanning and imagery for photorealistic 3D architectural documentation. In: Wang,
C.C. (ed.) Laser Scanning, Theory and Applications, pp. 414–430. InTech (2011)
11. Alsadik, B., Gerke, M., Vosselman, G., Daham, A., Jasim, L.: Minimal camera networks for
3D image based modeling of cultural heritage objects. Sensors 14(4), 5785–5804 (2014)
12. Themistocleous, K., Ioannides, M., Agapiou, A., Hadjimitsis, D.G.: The methodology of
documenting cultural heritage sites using photogrammetry, UAV, and 3D printing techniques:
the case study of Asinou Church in Cyprus. In: RSCy2015, pp. 953510-1–953510-7
International Society for Optics and Photonics (2015)
13. Alsadik, B., Gerke, M., Vosselman, G.: Efficient use of video for 3D modelling of cultural
heritage objects. ISPRS Ann. 2(3/W4), 1–8 (2015)
14. Kyriakaki, G., Doulamis, A., Doulamis, N., Ioannides, M., Makantasis, K., Protopapadakis,
E., Hadjiprocopis, A., Wenzel, K., Fritsch, D., Klein, M., Weinlinger, G.: 4D reconstruction
of tangible cultural heritage objects from web-retrieved images. Int. J. Herit. Digit. Era 3(2),
431–451 (2014)
15. Makantasis, K., Doulamis, A., Doulamis, N., Ioannides, M.: In the wild image retrieval and
clustering for 3D cultural heritage landmarks reconstruction. Multimed. Tools Appl. 75(7),
3593–3629 (2014)
16. Arias, P., Herraez, J., Lorenzo, H., Ordonez, C.: Control of structural problems in cultural
heritage monuments using close-range photogrammetry and computer methods. Comput.
Struct. 83(21), 1754–1766 (2007)
17. San José, J.I., Fernández Martín, J.J., Pérez Moneo, J.D., Finat, J., Martínez Rubio, J.:
Evaluation of structural damages from 3D laser scans. In: XXI CIPA International Symposium
(2007)
18. Alshawabkeh, Y., Bal’awi, F., Haala, N.: 3D digital documentation, assessment, and damage
quantification of the Al-Deir monument in the ancient city of Petra, Jordan. Conserv. Manage.
Archaeol. Sites 12(2), 124–145 (2010)
19. Crespo, C., Armesto, J., González-Aguilera, D., Arias, P.: Damage detection on historical
buildings using unsupervised classification techniques. ISPRS Arch. 38(Part 5) (2010)
20. Bhadrakom, B., Chaiyasarn, K.: As-built 3D modeling based on structure from motion for
deformation assessment of historical buildings. Int. J. GEOMATE 11(24), 2378–2384 (2016)
21. INACHUS, http://www.inachus.eu
22. Athanasiou, G., Amditis, A., Riviere, N., Makri, E., Bartzas, A., Anyfantis, A., Werner, R.,
Axelsson, D., di Girolamo, E., Balet, O., Schaap, M., Kerle, N., Bozabalian, N., Marafioti,
G., Berzosa, J., Gustafsson, A.: INACHUS: Integrated wide area situation awareness and
survivor localisation in search and rescue operations. In: GiT4NDM (2015)
23. Kerle, N., Fernandez Galarreta, J., Gerke, M.: Urban structural damage assessment with
oblique UAV imagery, object-based image analysis and semantic reasoning. In: 35th Asian
conference on remote sensing (2014)
24. Vetrivel, A., Gerke, M., Kerle, N., Vosselman, G.: Identification of damage in buildings based
on gaps in 3D point clouds from very high resolution oblique airborne images. ISPRS J.
Photogramm. Remote Sens. 105, 61–78 (2015)
25. RECONASS, http://www.reconass.eu
26. Gerke, M., Kerle, N.: Automatic structural seismic damage assessment with airborne oblique
Pictometry imagery. Photogramm. Eng. Remote Sens. 77(9), 885–898 (2011)
27. Verykokou, S., Ioannidis, C.: Metric exploitation of a single low oblique aerial image. In: FIG
Working Week (2015)
28. Verykokou, S., Ioannidis, C.: Exterior orientation estimation of oblique aerial imagery using
vanishing points. ISPRS Arch. 41(B3), 123–130 (2016)
29. Gerke, M.: Dense matching in high resolution oblique airborne images. ISPRS Arch. 38(Part
3/W4), 77–82 (2009)
30. Lambrou, E., Pantazis, G.: Geometric documentation of structures after an earthquake. In:
FIG Working Week (2004)
31. Verykokou, S., Ioannidis, C., Kontogianni, G.: 3D visualization via augmented reality: the
case of the Middle Stoa in the Ancient Agora of Athens. In: Ioannides, M., Magnenat-
Thalmann, N., Fink, E., Žarnić, R., Yen, A.-Y., Quak, E. (eds.) EuroMed 2014. LNCS, vol.
8740, pp. 279–289. Springer, Heidelberg (2014). doi:10.1007/978-3-319-13695-0_27
32. Ioannidis, C., Verykokou, S., Soile, S., Potsiou, C.: 5D multi-purpose land information
system. In: Biljecki, G., Tourre, V. (eds.) Eurographics Workshop on Urban Data Modelling
and Visualisation, pp. 19–24. The Eurographics Association (2015)
33. Ioannidis, C., Potsiou, C., Soile, S., Verykokou, S., Mourafetis, G., Doulamis, N.: Technical
aspects for the creation of a multi-dimensional land information system. ISPRS Arch. 41(B2),
115–122 (2016)
34. Doulamis, A., Doulamis, N., Ioannidis, C., Chrysouli, C., Grammalidis, N., Dimitropoulos,
K., Potsiou, C., Ioannides, M.: 5D modelling: an efficient approach for creating spatiotemporal
predictive 3D maps of large-scale cultural resources. ISPRS Ann. 2(5/W3), 61–68 (2015)
35. Agisoft, http://www.agisoft.com
36. Verykokou, S., Doulamis, A., Athanasiou, G., Ioannidis, C., Amditis, A.: UAV-based 3D
modelling of disaster scenes for urban search and rescue. In: IEEE IST (2016)
37. Lowe, D.G.: Distinctive image features from scale-invariant keypoints. Int. J. Comput. Vision
60(2), 91–110 (2004)
Low Cost Technique for Accurate Geometric
Documentation of Complex Monuments by Non-experts
Charalabos Ioannidis ✉ , Sofia Soile, and Styliani Verykokou

( )
School of Rural and Surveying Engineering,

National Technical University of Athens, Athens, Greece
{cioannid,ssoile}@survey.ntua.gr, [email protected]
Abstract. This paper proposes the technique of stereo-orthoimage for reliable

and accurate identification and digitization of complex features of cultural
heritage (CH) monuments (e.g., edges, outlines, damages, holes, cracks) in the
context of large-scale geometric and damage documentation. It is a low cost
technique, which can be implemented by non-experts (architects, archaeologists,
etc.). A developed plugin, named OrthoSteroMate (OSM), for the open-source
GIS system QGIS that implements the stereo-orthoimage technique is presented.
It introduces stereo-orthoimages in GIS environments, as complements to
conventional orthoimages, allowing better interpretation of the details of built CH
and enabling more accurate digitization, taking advantage of stereoscopic obser‐
vation when no special equipment or photogrammetric knowledge are needed.
The application of the plugin along with QGIS tools for the restitution of two CH
monuments were made, yielding satisfying results and proving the applicability
of the proposed low-cost method for complex CH documentation.
Keywords: Anaglyph image · Artificial x-parallax · Geometric documentation ·

Orthoimage · Stereo-mate · Stereo-orthoimage
1 Introduction
The geometric documentation of monuments requires the acquisition, processing,

presentation and recording of the necessary data for the determination of their position
and their form, shape and size in the 3D space at a particular moment in time [1]. It
forms the necessary background not only for the studies of their past, but also for their
preservation and protection as well as the studies of their future. The usual products for
the geometric documentation of monuments consist of orthoimages, 2D vector drawings
and 3D models having metric properties [2]. Orthoimages, generally created by photo‐
grammetrists, are extensively used as a background by professionals of various special‐
ties, like architects and archaeologists, for the digitization of damages of monuments,
broken edges and cracks in addition to the typical digitization of monuments for the
production of 2D vector drawings. Cultural damage digitization is a prerequisite for the
detection of deformations and cracks [3] and vector drawings indicating damages may
be used as inputs in a monument damage information system [4] as well as in GIS
systems, supporting strategic planning of possible conservation interventions for the

DOI: 10.1007/978-3-319-48496-9_6
64 C. Ioannidis et al.
protection, management and sustainable development of built cultural heritage (CH) [5].
However, the fact that the surfaces of monuments are usually complex, with intense
anaglyph, makes the 2D digitization process via orthoimagery a rather difficult task,
which may end up in mistakes. Thus, for complex monuments with irregular edges and
damages, the possibility for 3D stereoscopic observation would significantly facilitate
the digitization process and contribute to more accurate products.
This paper proposes the technique of stereo-orthoimage for use in the digitization
process of monuments with complex features and damages, as it combines the advan‐
tages of both the orthoimage and 3D observation, enabling accurate identification of
complex features. In cases of monuments with intense anaglyph and various damages,
the stereoscopic observation could allow efficient monitoring of the digitized edges and
detection of the digitization errors, which can be easily corrected through stereoscopic
measurements. The stereo-orthoimage based digitization does not require specific
photogrammetric knowledge and specialized equipment, thus being a low-cost tech‐
nique that can be adopted even by non-experts.
The concept of stereo-orthoimage was introduced at the end of the 1960s [6] with
applications in the areas of Cadastre and medium scale mapping. The theory of stereo-
orthoimages and their applications, concerning the updating of maps and cadastral
diagrams, are presented in [7]. The Ortho Stereo Mate (OSM) technique, that is, the
artificial creation of a “partner” (mate) orthoimage in order to enable stereoscopic
observation, was proposed in 1976 [8, 9] for use in developing countries, for easy
mapping and accurate photo interpretation of large areas, achieving a significant
improvement in the quality of the final products. However, the limitations of this tech‐
nique, that is, the fact that is uses a derivative product of original images (i.e., orthoim‐
ages) and its lack of adaptation in the instruments and techniques of the analog and
analytical photogrammetry, led to the reduction of its use during the 1980s and 1990s.
During the last 20 years, the interest towards the stereo-orthoimage technique was
revitalized using digital methods [10–12] in combination with GIS systems [13–16], as
the latter ones have a large number of users of various scientific fields and disciplines,
for whom the orthoimages are products with high metric and photointerpretation value.
However, current research is more focused in airborne applications and the potential of
stereo-orthoimages in close-range applications with architectural or archaeological
interest has not been fully exploited yet. In this context, the purpose of this paper is the
introduction of stereo-orthoimages in GIS environments as complements to conven‐
tional orthoimages to allow better interpretation of the details of complex and damaged
CH monuments even by non-experts. Thus, a free plugin for an open-source GIS, QGIS,
was developed; it implements the stereo-orthoimage technique enabling stereoscopic
observation with the use of color-coded anaglyph glasses.
2 Geometric Documentation of Monuments Using 2D and 3D

Techniques and Products
The geometric documentation of CH monuments may be performed via simple topo‐

metric methods, if the complexity of the monument allows it, or, usually, via the
Low Cost Technique for Accurate Geometric Documentation 65
combination of surveying and photogrammetric methods. The latter ones rely on direct
measurements, both in the field and using images, for the creation of 2D products, like
orthoimages and vector drawings (e.g., horizontal sections, cross-sections), and 3D
outputs, like point clouds and textured meshes. The photogrammetric methods for
geometric documentation of monuments may be categorized into monoscopic (e.g.,
orthoimage, projective rectification for flat surfaces), stereoscopic (e.g., stereo-restitu‐
tion) and multi-image (e.g., image-based 3D modelling) [2]. The use of terrestrial laser
scanners (TLS), in most cases in combination with photogrammetric and surveying
techniques, may yield impressive results on 3D documentation of complex monuments
[17]. However, laser scanning is a rather costly solution that may be replaced by image-
based 3D point cloud generation via dense image matching techniques [18].
The introduction of contemporary techniques and tools (TLS, computer vision algo‐
rithms, multi-view systems, etc.) in CH documentation, allows the collection of large
volumes of 3D data (images and point clouds) and the creation of 3D models of monu‐
ments using automated procedures. On the other hand, architects and other CH profes‐
sionals continue to use mainly 2D vector drawings on their studies, as the 3D viewers
do not have yet the flexibility or the tools needed for these studies, and the 3D (image)
processing software is expensive and complex in use. So, the orthoimages which can be
produced easily from the 3D models, are the best background for the 2D digitization of
monuments by professionals without particular photogrammetric knowledge. However,
as they do not allow stereoscopic observation, the resulting vector drawings may contain
digitization errors and inaccuracies, especially in cases of complex monuments with
damages and intense relief, where the identification of edges is a challenging task, mainly
for non-experts. In such cases, stereo-orthoimages would be low-cost photogrammetric
products that do not require specialized equipment and permit accurate documentation
thanks to stereoscopic observation.
3 The Proposed Method
The proposed method for the geometric documentation of complex and damaged monu‐
ments requires the performance of three fundamental steps: (a) Digital Surface Model
(DSM) creation, (b) orthoimage or orthomosaic creation and (c) stereo-orthoimage
generation. The input data consist of multiple overlapping terrestrial and UAV imagery
and ground control points (GCPs) acquired via topographic methods.
3.1 DSM and Orthoimage or Orthomosaic Generation
The DSM of a monument is used both for the orthoimage or orthomosaic creation and
the generation of the stereo-orthoimage. The image-based creation of a DSM requires
the implementation of two fundamental techniques, that is, Structure from Motion (SfM)
and Dense Image Matching (DIM). SfM is a low-cost technique that refers to the process
of recovering the 3D structure and the camera poses (i.e., the camera exterior orientation)
from a 2D image sequence [19]. The extraction of feature points in each image, followed
by the matching of the descriptors of the feature points, using the criterion of a minimum
distance measure and techniques for outlier removal, are the first steps of the SfM
process. In the case of non-calibrated cameras, a projective reconstruction is applied and
is transformed into a metric one via auto-calibration. In the case of calibrated cameras,
a metric reconstruction is directly obtained. The final step of the SfM process is usually
a bundle adjustment. The georeferencing of the SfM point cloud and the camera projec‐
tion centers is generally performed via the computation of the 3D similarity transfor‐
mation between the arbitrary SfM coordinate system and the world reference system via
GCPs. DIM using the camera external parameters estimated by SfM is applied for the
creation of the DSM of the monument of interest. Several DIM algorithms may be
implemented using either a stereo or a multi-view approach, according to which, the
DSM may be generated at a resolution which corresponds to the ground sampling
distance of the original images [20]. A review and analysis of DIM algorithms can be
found in [21].
Finally, the creation of an orthoimage or othomosaic of a monument or part of a
monument (e.g., front or top view) is accomplished through differential rectification. It
requires the knowledge of the camera interior orientation, the exterior orientation of the
image or images used for the generation of the orthoimage or mosaic respectively and
the DSM of the region depicted in the image(s). A commercial (e.g., Agisoft PhotoScan,
Pix4Dmapper, iWitness, ContextCapture, SURE) or open-source (e.g., MicMac, VSfM)
software package may be used for the creation of a DSM and orthoimage or orthomosaic
of CH monuments for being used in the proposed stereo-orthoimage technique via the
developed QGIS plugin.
3.2 The Stereo-Orthoimage Technique

The basic concept of the stereo-orthoimage technique is the creation of a stereo-pair
consisting of orthoimages, which enables stereoscopic observation. The ortho stereo-
pair consists of an orthoimage, which may have been produced from the left or the right
image of the pair of original images, and the Ortho Stereo Mate (OSM), that is, the
“partner” (mate) orthoimage, which has been artificially created in order to allow ster‐
eoscopic observation. The generation of the OSM is achieved by introducing artificial
x-parallax in each pixel of the orthoimage. The stereo-orthoimage technique is illustrated
in Fig. 1.
Artificial Parallax Calculation. The artificial x-parallaxes include information about

the third dimension and their introduction to the orthoimage creates a new image; the
latter along with the orthoimage constitute the ortho stereo pair. The artificial x-parallax
for every pixel is calculated as a function of its third dimension, derived from the DSM
of the monument, which has also been used for the orthoimage production. Several
methods for calculating the artificial x-parallax have been proposed (e.g., the parallel
projection method, the logarithmic method, the non-parallel projection method) [16].
The non-parallel projection method is proposed in this paper, because the calculated
artificial x-parallax is, by definition, of the same size as the “original” one. According
to this method, the x-parallax is calculated by Eq. (1):
Fig. 1. The stereo-orthoimage technique
B⋅Z
P= (1)
H
where B is the base of the stereo pair of orthoimages, Z is the third dimension of the
point, the x-parallax (P) of which is calculated, and H is the “flying” height, that is, the
distance between the camera and the reference level of the monument.
Ortho Stereo Mate Generation. The initial data required for the creation of the OSM
include an orthoimage, its corresponding DSM, the base of the generated stereo pair of
imagery and the “flying” height. The horizontal ground coordinates of every pixel of
the orthoimage are calculated using the georeferencing information stored in the world
file of the orthoimage. The elevation of every pixel is calculated through the bilinear
interpolation in the DSM. The artificial x-parallax for every pixel is computed according
to Eq. (1). The abscissas of the OSM are computed via the calculation of the new posi‐
tions where the geodetic X coordinates of the orthoimage have to be moved, because of
the introduction of the artificial x-parallax, according to Eq. (2):
XOSM = Xortho − P. (2)
The X coordinates of the OSM along with the ordinates of the orthoimage and their
corresponding Z values form a 3D point cloud. The OSM has the same pixel size as the
orthoimage, in order to allow stereoscopic observation. The number of rows (height) of
this image is the same as the number of rows of the orthoimage. However, the number
of columns (width) of these images, differs, because of the introduction of the artificial
x-parallax. The number of columns of the OSM is given by Eq. (3):
XOSM,max − XOSM,min
colsOSM = . (3)
cellortho
Only the blue and green channel of the orthoimage are used for the OSM generation.
Points having bigger Z value than the Z value of their neighboring ones are displayed
to the left of their neighbors with lower elevation and smaller abscissa. Thus, the geodetic
X coordinates of the cells of the OSM are stored in ascending order in every row and
their initial position in the row (i.e., the number of their initial column) is also stored.
The cells with smaller initial column number than the one after the ordering are deleted
and black cells cover the OSM in these areas. The color value of each cell of the OSM
is determined according to the following assumptions [16]:
– if the pixel of the OSM fully coincides with a “shifted” cell of the orthoimage or
partially coincides with a “shifted” cell of the orthoimage and the rest of it lies in an
area without cells (color values), the color value of the corresponding pixel of the
orthoimage is given to the pixel of the OSM;
– if the pixel of the OSM covers two or more “shifted” cells of the orthoimage, the
color value of the pixel having the bigger Z value between the two nearest ones is
given to the pixel of the OSM;
– if the pixel of the OSM lies in a region that is not covered by any “shifted” cell of
the orthoimage, the black color (zero tone) is given to it.
The stereo-orthoimage, that is, the anaglyph image, is generated via the composition
of the red channel of the orthoimage and the blue and green channel of the OSM; it
enables stereo observation, provided that the operator is equipped with special color-
coded anaglyph glasses, which are usually red and cyan.
3.3 Digitization of Complex Monuments
OSM Plugin for QGIS. QGIS, developed by the Open Source Geospatial Foundation
(OSGeo), is a free and open source GIS used by scientists of various disciplines. A QGIS
plugin, named OrthoStereoMate (OSM), was developed, implementing the stereo-
orthoimage technique (Fig. 2); it expands the functions of QGIS and covers the
constraints of current software solutions regarding stereoscopic observation and digiti‐
zation. For instance, ArcScene enables stereoscopic observation and digitization via the
creation of an anaglyph image; however, some tests conducted revealed a practical
weakness for points with high relief, as the allowable range of motion to eliminate the
x-parallax is not sufficient. Also, although stereoscopic observation is possible through
ERDAS Imagine, via the creation of a color anaglyph simulation of the terrain in three
dimensions, the digitization process is not feasible.
The OSM plugin was developed in the Python programming language and its graph‐
ical user interface (GUI) was developed via the Qt cross-platform application frame‐
work, through Python bindings for Qt (PyQt) using the integrated development envi‐
ronment Qt Creator. The fact that the whole QGIS code depends on Qt libraries allows
the seamless integration of PyQGIS (Python bindings for QGIS) with PyQt. The neces‐
sary files of the Python plugin were created according to the directory structure defined
Fig. 2. OrthoSteroMate plugin GUI
by the QGIS Project. The OSM plugin requires the addition of two raster layers in QGIS
environment: the orthoimage and the DSM. The base between the orthoimage and the
OSM image, and the distance between the camera and the reference level of the monu‐
ment have to be specified in the corresponding input widgets of the plugin. Then, the
OSM can be created, via the corresponding button, and is automatically added as a raster
layer in QGIS. The next step is the creation of the anaglyph image (stereo-orthoimage)
via the corresponding button; this image enables 3D observation, with the use of
anaglyph glasses. Similarly, the anaglyph image is added as a raster layer. The x-parallax
correction in the anaglyph image may be accomplished via the appropriate buttons,
which move the right orthoimage 1, 5 or 10 pixels to the left or right, and recreate the
anaglyph image. Alternatively, if the predefined number of pixels (1, 5 and 10) does not
meet the requirements of the application, the x-parallax correction is accomplished by
moving the horizontal slider to the target number of pixels by which the right orthoimage
has to be moved and recreating the anaglyph image via the appropriate button. Finally,
the user may create or load vector layers in QGIS and start digitizing using the available
QGIS tools. At any time, the x-parallax can be corrected locally in the relief image, via
the OSM plugin.
Digitization of Complex Monuments Using the OSM Plugin and QGIS Tools. The
stereo-orthoimage generated via the OSM plugin may be used:
– For stereoscopic observation and checking of the digitization errors arisen due to 2D
recording of edges, outlines, damages, cracks, holes, cavities, etc. using conventional
orthoimages and stereoscopic re-digitization in areas with errors;
– For stereoscopic digitization of features non-easily identifiable in 2D orthoimages;
– For complete stereoscopic digitization of complex and damaged monuments without
the use of 2D orthoimagery.
Figure 3 illustrates the proposed stereoscopic digitization procedure using QGIS and
the OSM plugin. Stereo orthoimages generated via the plugin may be used for the crea‐
tion of detailed 2D vector drawings of damaged monuments, which consist of irregular
or unclear edges that cannot be easily identified using conventional orthoimages. The
proposed procedure results into better, more reliable and accurate recording of outlines,
edges, damages and cracks in comparison with the use of a single (2D) orthoimage for
digitization, because of stereoscopic observation. The precision of the measurements
remains the same as the precision of the orthoimage but the errors and inaccuracies
because of wrong interpretation of damaged edges are eliminated via the proposed tech‐
nique. Whereas the generation of the initial data (i.e., DSM and orthoimage) has to be
carried out by photogrammetrists, the restitution via the presented stereoscopic digiti‐
zation process using the plugin may be performed even by non-experts.
Fig. 3. Damage digitization using the OSM plugin and QGIS tools
4 Applications
The OSM plugin was used along with QGIS tools for the geometric documentation of
a variety of CH monuments with different size, construction era, style, from all over
Greece; the common characteristic of all these monuments was the significant damage
of their materials. In the following, two of the monuments are described as case studies.
4.1 Data Description and Processing
The first case study concerns the ruins of a stone ancient tower, which was a part of an
ancient wall in Attica, Greece. The remaining part of the tower has an intense relief of
2.5 m, with stones which have irregular edges and suffer from corrosion and material
damages. This monument is located in the internal area of a contemporary building,
structural elements of which (such as walls and columns) are in contact with the stones
of the ancient wall. Terrestrial images were acquired by a Canon EOS 5 Mark II camera
with a 21.1-megapixel sensor and a prime lens of 24 mm. They were captured from
distances ranging from 3 to 6 m from several perspectives, with large overlap between
them. GCPs were also measured. Τhe DSM and the orthoimage of the top view of the
tower corresponding to an area of 70 m2 were created through Agisoft PhotoScan,
corresponding to a pixel size of 7.5 mm and a cell size of 3 mm respectively (Fig. 4 left).
The second application concerns the restitution of the crepidoma (a platform
constructed of three steps on which the superstructure of the building is erected) of an
Fig. 4. Orthoimages of the top view of the ancient tower (left) and the facade of the crepidoma
of the temple (right)
ancient temple in Athens, Greece, made of marble, having an anaglyph variation of

80 cm. Several damages, corrosion and irregular edges are present in this part of the
monument. Terrestrial images were acquired by a Canon EOS 5 Mark II camera with a
lens of 24 mm. The images were captured from distances ranging from 5 to 7 m from
several perspectives and correspond to 80 % overlap. GCPs were also measured in the
field. The DSM and the orthoimage of the front view of the crepidoma were created via
Agisoft PhotoScan, corresponding to a pixel size of 3 mm and a cell size of 3 mm
respectively (Fig. 4 right).
4.2 Digitization Processes
For each case study, the orthoimage and the DSM were inserted into QGIS. The distance
(H) between the camera and the reference level of the monument, and the base (B) of
the stereo-pair of orthoimages were specified using the OSM plugin. Ratios B/H ranging
from 0.15 to 0.2 were used. Figure 5 top shows a view of the 3D model of the tower and
the stereo-anaglyph developed via the OSM plugin; the extensive damages of the stones,
especially in the front façade, that make the digitization via a 2D orthoimage difficult,
are shown here. Figure 5 bottom shows the produced stereo-anaglyph of part of the
crepidoma, which illustrates the intense anaglyph, the cracks and the missing material
that creates rough and uneven edges, which cannot be identified correctly or at all
through mono-image digitization.
For comparison reasons, for each case study two independent detailed digitizations
were performed; the first one was carried out in AutoCAD environment using the 2D
conventional orthoimages and the second one was performed through 3D observation
and digitization of stereo-orthoimages using the QGIS tools and the OSM plugin for x-
parallax correction. Figure 6 shows the environment where the digitization process has
been performed using the stereo-orthoimages of the ancient tower.
The digitization results derived from both techniques (using orthoimages and stereo-
orthoimages) were superimposed on one another and comparative studies were
performed. Most differences were observed in areas with strong elevation variations and
particularly in damaged and irregular edges. The restitution results derived from stereo-
digitization in QGIS – OSM environment are more reliable than the products generated
through mono-digitization of orthoimages and include more details.
Fig. 5. 3D model and stereo anaglyph of the ancient tower (top); stereo-anaglyph of the
crepidoma of the temple (bottom)
The products of the mono-digitization process include outliers, which are due to the
difficulty of edge recognition; thus, either edges are omitted (successive stones are not
distinguished from each other or the edges that are created due to a broken stone or
missing material are not digitized) or non-existent false edges are added (e.g., there is a
sense that an edge is there due to lighting difference). Figure 7 illustrates the digitization
outputs of both techniques superimposed on one another; at the tower less edges than
the existing number have been digitized using the mono-digitization process, because
the stones suffer from extended corrosion (Fig. 7 top), while at the crepidoma many
details of the stones have not been digitized (Fig. 7 bottom).
Fig. 6. Stereo-digitization using QGIS tools and the OSM plugin

Fig. 7. 2D digitization (blue color) and stereo-digitization (magenta color) outputs superimposed
on one another (Color figure online)
One weakness of the stereo-digitization process is the fact that additional time is
needed for the correction of the x-parallax. What is more, the restitution through mono-
digitization is easier to be performed by non-experts. However, the option to combine
the two technical methods for digitization in QGIS environment leads to a minimization
of such weaknesses. Thus, in areas where the edges are unclear, 2D digitization may be
applied via the user-friendly tools of QGIS and the OSM plugin may be used at the
remaining areas or just for checking the correctness of digitization.
5 Conclusions
This paper proposes the technique of stereo-orthoimage for use in the digitization of
complex monuments with damages and irregular edges. This low-cost technique enables
accurate identification of complex features and eliminates digitization errors and inac‐
curacies arising from 2D observation of conventional orthoimages, which are the most
common product of geometric documentation used as a background by architects and
archaeologists. A plugin for the QGIS (an open source software) was developed and
used along with QGIS tools for stereo-restitution purposes; there is no need for any
expensive hardware or software component. A comparative study on two CH monu‐
ments was performed, proving that the proposed technique improves the reliability,
accuracy and completeness of the digitization products in comparison with the use of
simple 2D orthoimages, in areas with intense relief, cracks, holes and damaged edges.
References
1. United Nations Educational, Scientific and Cultural Organization: Photogrammetry applied

to the survey of Historic Monuments, of Sites and to Archaeology. UNESCO Editions (1972)
2. Georgopoulos, A., Ioannidis, C.: Photogrammetric and surveying methods for the geometric
recording of archaeological monuments. In: FIG Working Week (2004)
3. Cheng, H.M.: The workflows of 3D digitizing heritage monuments. In: Munoz, J.A. (ed.)
Laser Scanner Technology. InTech Open Access Publisher (2012)
4. Cacciotti, R., Valach, J., Kuneš, P., Cernanský, M., Blaško, M., Kremen, P.: Monument
damage information system (MONDIS): an ontological approach to cultural heritage
documentation. ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci. 2(5), 55–60 (2013)
5. Moropoulou, A., Labropoulos, K.C., Delegou, E.T., Karoglou, M., Bakolas, A.: Non-
destructive techniques as a tool for the protection of built cultural heritage. Constr. Build.
Mater. 48, 1222–1239 (2013)
6. Collins, S.: The stereorthophoto pair. Photogramm. Eng. 38(12), 1195–1202 (1972)
7. Blachut, T.J.: Mapping and photointerpretation system based on stereo-orthophotos. Doctoral
dissertation, Institute of Geodesy and Photogrammetry, ETH Zürich (1971)
8. Blachut, T.J.: Results of the International orthophoto experiment 1972-1976. Photogramm.
Eng. Rem. Sens. 42(12), 1483–1498 (1976)
9. Kraus, K., Otepka, G., Lottsc, J., Haitzmann, H.: Digitally controlled production of
orthophotos and stereo-orthophotos. Photogramm. Eng. Rem. Sens. 45(10), 1353–1362
(1979)
10. Damen, M.C.J., Smit, G.S., Verstappen, H.T.: Remote sensing for resources development and
environmental management. In: 7th International Symposium on Remote Sensing for
Resources Development and Environmental Management (1986)
11. Domik, G., Leberl, F., Cimino, J.: Multiple incidence angle SIR-B experiment οver Argentina:
generation of secondary image products. IEEE Trans. Geosci. Remote Sens. 24(4), 492–497
(1986)
12. Chang, H., Yu, K., Joo, H., Kim, Y., Kim, H., Choi, J., Han, D.Y., Eo, Y.D.: Stereo-mate
generation of high resolution satellite imagery using a parallel projection model. Can. J.
Remote Sens. 34(2), 57–67 (2008)
13. Baltsavias, E.P.: Integration of ortho-images in GIS. In: Optical Engineering and Photonics
in Aerospace Sensing, pp. 314–324. SPIE (1993)
14. Baltsavias, E.P.: Digital ortho-images – a powerful tool for the extraction of spatial-and geo-
information. ISPRS J. Photogramm. Remote Sens. 51, 63–67 (1996)
15. Pyka, K., Slota, M., Twardowsk, I.M.: Usage of stereo orthoimage in GIS: old concept,
modern solution. In: The International Archives of the Photogrammetry, Remote Sensing and
Spatial Information Sciences, vol. XXXIX-B4, pp. 379–383 (2012)
16. Soile, S., Verykokou, S., Ioannidis, C.: Stereo-orthoimage as a GIS tool for reliable restitution
of Cultural Monuments. Int. J. Heritage Digit. Era 4(3–4), 275–294 (2015)
17. Remondino, F.: Heritage recording and 3D modeling with photogrammetry and 3D scanning.
Remote Sens. 3(6), 1104–1138 (2011)
18. Kersten, Thomas, P., Lindstaedt, M.: Image-based low-cost systems for automatic 3D
recording and modelling of archaeological finds and objects. In: Ioannides, M., Fritsch, D.,
Leissner, J., Davies, R., Remondino, F., Caffo, R. (eds.) EuroMed 2012. LNCS, vol. 7616,
pp. 1–10. Springer, Heidelberg (2012). doi:10.1007/978-3-642-34234-9_1
19. Remondino, F., Pizzo, S., Kersten, T.P., Troisi, S.: Low-cost and open-source solutions for
automated image orientation – a critical overview. In: Ioannides, M., Fritsch, D., Leissner, J.,
Davies, R., Remondino, F., Caffo, R. (eds.) EuroMed 2012. LNCS, vol. 7616, pp. 40–54.
Springer, Heidelberg (2012). doi:10.1007/978-3-642-34234-9_5
20. Haala, N: The landscape of dense image matching algorithms. In: Fritsch, D. (ed.)
Photogrammetric Week 2013, pp. 271–284. Wichmann/VDE Verlag, Belin & Offenbach
(2013)
21. Remondino, F., Spera, M.G., Nocerino, E., Menna, F., Nex, F., Gonizzi-Barsanti, S: Dense
image matching: comparisons and analyses. In: Digital Heritage International Congress, vol.
1, pp. 47–54 (2013)
Full Paper: Heritage Building
Information Models (HBIM)
Implementation of Scan-to-BIM and FEM
for the Documentation and Analysis
of Heritage Timber Roof Structures
Maarten Bassier1(&), George Hadjidemetriou1, Maarten Vergauwen1,

Nathalie Van Roy2, and Els Verstrynge2
1
Department of Civil Engineering, TC Construction – Geomatics, KU Leuven,
Leuven, Belgium
{maarten.bassier,maarten.vergauwen}@kuleuven.be,
[email protected]
2
Building Materials and Building Technology Division, Department of Civil
Engineering, KU Leuven, Leuven, Belgium
{nathalie.vanroy,els.verstrynge}@kuleuven.be
http://www.kuleuven.be
Abstract. Current heritage analysis applications and documentation techniques

for timber roof structures rely on manual measurements to provide the spatial
data. Major simplifications are made to document these structures efficiently.
However, these simplified geometric models provide less reliable results.
Therefore, the need exists for more realistic models. Additionally, the
exchangeability of information between varying parties is paramount. Hence,
the construction elements should be defined in a Building Information Model
(BIM). This allows users to reuse the model, allowing the distribution of
information throughout the project. The goal of our research is to create a
realistic BIM model of a complex heritage roof structure employing dense point
clouds. The comparison of our complex geometric model to a traditional
wire-frame model proves that our approach provides more reliable results in
terms of geometry and structural behaviour. Our work covers the acquisition, the
modelling and the structural analysis of timber roof structures.
Keywords: Cultural heritage BIM Structural analysis Modelling Timber

roof structures
1 Introduction
Built heritage is an integral part of our history and its conservation is important. To
preserve these assets, qualitative as well as quantitative information about the structure
is needed throughout the conservation process. The acquired data is used for the
documentation, analysis and exchange of the cultural heritage. Current heritage projects
are struggling with such data management. For instance, the spatial measurements of
heritage monuments prove problematic due to the complexity of these structures. This
is particularly true for timber roof structures. Subsequently, the metric data is used to

DOI: 10.1007/978-3-319-48496-9_7
80 M. Bassier et al.
create geometric models. The quality of these models have a major impact on the
structural analysis of the structure. Less realistic geometry provide less accurate results.
Another major factor in the data management process is data homogeneity. The
information at consecutive stages of the process should be accessible by the varying
stakeholders. However, due to software incompatibility, the data exchange between
parties is flawed or non-existent. As a result, specialists have to revisit the site and
reacquire the scene over and over again.
The goal of our research is to provide the heritage industry with a workflow that
will allow for more comprehensive documentation, the creation of more realistic
geometric models and the improvement of interoperability between parties. This paper
is organised as follows. Subsections 1.1, 1.2 and 1.3 give detailed information about
the state of the art of heritage documentation, analysis and data exchange. Section 2
elaborates on the methodology. Sections 3 and 4 respectively discuss the test design
and results. In Sect. 5, the conclusions are presented.
1.1 Data Acquisition

Traditionally, hand measurements and visual inspections are employed to document
cultural heritage. Such measurements, for instance, serve as a basis to create geometric
models used in structural analysis. However, the sparsity and inaccuracy of hand
measurements only allow for the creation of simplistic models. Denser and more
accurate data is required to create more realistic geometric models.
An increasingly popular data acquisition tool is Terrestrial Laser Scanning (TLS).
The static scanning instrument is capable of capturing accurate 3D point measurements
of an entire scene in a matter of minutes. The result is a geometric point cloud con-
taining tens of millions of points with high accuracy. An example of a scanned roof
structure is shown in Fig. 1.
Terrestrial Laser Scanning and point cloud data have been used in several heritage
projects [1]. In addition to TLS data, photogrammetric point clouds are also used [2]. In
our research, we prefer to use TLS data over photogrammetric point clouds because of
Fig. 1. Overview heritage roof structure: image (left), point cloud (right).
Implementation of Scan-to-BIM and FEM 81
the size of the heritage site. Research indicates that Terrestrial Laser Scanning, while
being more expensive, scales better with the size of the project in terms of accuracy and
density [3].
1.2 Geometric Models for Structural Analysis

The analysis of the structural behaviour is crucial to the assessment and preservation of
built heritage [4]. The aim of this analysis is appropriate damage therapy and
strengthening measures. Currently, basic wire-frame models based on hand measure-
ments are employed for the geometric representation. These simplistic models are error
prone and provide only a rough approximation of the real conditions. Therefore, the
analysis is less reliable.
Several researchers have proposed more complex geometric models. A promising
approach is the use of solid models [5]. Castellazzi presented such models based on
voxel elements [6]. Their semi-automated procedure has shown promising results
compared to traditional CAD-based models. Also constructive solid geometry
(CSG) has been considered [7]. They state that solid approaches are to be preferred
over wire-frame models for mass calculations and structural behaviour. In our research,
we also employ a volume based representation of the geometry. This allows for a more
realistic representation of the structural elements.
1.3 Data Exchange

Data heterogeneity is a major obstacle in the heritage industry. During a project, data is
stored in different software specific formats conform varying data models. The data is
not interoperable with other applications. To facilitate the data transfer throughout the
conservation process, the need exist for a uniform data model and an exchangeable
format. An innovative technology is Building Information Models (BIM). These
models constitute a digital representation of a structure with all its metric, non-metric
and semantic data. The data is stored in a format called Industry Foundation Classes
(IFC), an open standard for the exchange of BIM information between different soft-
ware. In our research, we employ IFC to store the geometry of structural elements.
Research has been performed on the integration of BIM in the cultural heritage
industry [8–10]. In addition, heritage BIM library projects such as HBIM provide a set
of reusable heritage objects [11–13]. BIM has also been used as a basis for the
structural analysis of heritage buildings [14]. This field is still ongoing research.
2 Methodology
In this paper we propose an approach to create realistic BIM objects of structural

elements based on point cloud data and apply these BIM objects as a base for structural
analysis. More specifically, we create a model for the evaluation of a heritage timber
roof structure. Our approach consists of two phases. The Scan to BIM phase covers the
data acquisition, modelling of the elements and their representation in a BIM
Fig. 2. Workflow for Scan-to-BIM and structural analysis.
environment. The structural analysis phase covers the exchange of the model to a
Structural Analysis software and the evaluation of the model’s behaviour. An overview
of the intermediate stages is depicted in Fig. 2. Our model is compared to a traditional
geometric wire-frame model, which functions as a benchmark.
2.1 Scan to BIM

For the data acquisition, we employ a static terrestrial laser scanner to acquire a dense
3D point cloud of the structure. Leica Cyclone is utilized to process the data. First,
cloud-to-cloud constraints are defined for the coarse alignment of each scan. Subse-
quently, an Iterative Closest Point (ICP) Algorithm computes a fine alignment between
the data sets.
After the registration, the unified point cloud is exported to 3D Reshaper for further
modelling (Fig. 3a). In this software, watertight meshes are computed for each element
based on the point cloud. The software employs a coarse to fine meshing approach. The
coarse meshing computes a rough shape representation to acquire the initial shape. The
fine meshing refines the initial geometry to better approximate the point cloud. Both
stages are an implementation of a Poisson meshing algorithm [15] (Fig. 3b). Second,
the mesh is interpolated to create a watertight mesh (Fig. 3c). Third, the mesh is
segmented per element and closed (Fig. 3d). For structural analysis, it is imperative that
the connections between elements are consistent so they are able to transfer forces
correctly. At this stage, the data is defined by a non-parametric representation. While
this representation is adequate for simple evaluations, applications that interact with the
geometry prefer parametric data representations.
The closed meshes serve as a basis for the parametric reconstruction of the beams.
The objects are imported in the SolidWorks software (Fig. 4a). The built-in ScanTo3D
function allows for the semi-automated reverse engineering of closed meshes. This
function operates as follows. First, the mesh is represented by a set of surface patches to
approximate the exact geometry (Fig. 4b). The user can influence the settings of the
reconstruction algorithm to alter the level of detail. This has a direct impact on the
Fig. 3. Overview stages of the modelling process in 3D Reshaper: point cloud (a), Poisson mesh (b),
watertight mesh (c), segmented mesh (d).
number of generated patches. Next, the user is able to adjust topological errors present
in the patch representation (Fig. 4c). Finally, the software calculates the parametric
volumetric representation of the given patches (Fig. 4d). After the modelling, the
objects are exported to the IFC 2 × 3 format, which is exchangeable with BIM soft-
ware such as Revit.
2.2 Structural Analysis

A Finite Element Model (FEM) is defined for the structural analysis. In addition to the
geometry, this model encompasses loads, material properties and joint definitions.
These parameters are derived from varying standards [16–20]. The Finite Element
Analysis (FEA) is performed in the ANSYS software. Quadratic triangular volume
elements are used to represent the geometry [21]. A 3D solid analysis method is chosen
to perform the evaluation. For comparison purposes, a wire-frame model of the roof
structure is created from manual measurements. The same method is employed to
Fig. 4. Overview stages of the modelling process in SolidWorks: triangulated mesh (a), surface
patches on normal LOD (b), surface patches on high LOD (c), solid parametric model (d).
define the FEM for this model. The wire-frame model will function as a reference in the
evaluation in Sect. 4.
3 Test Design
Our approach is evaluated using realistic conditions. The test site is a small castle
located in the city of Mechelen, Belgium. The structure is nearly 400 years old and is
heavily deteriorated [22]. The roof itself shows major deflections and damaged
structural members. The supporting elements consist of oak beams. The structure was
scanned employing a FARO Focus 3D X330. A total of 54 scans were acquired to map
the building, resulting in a point cloud of approximately 540 million points. The roof
itself was acquired with 16 scans. A scanning resolution of 12.5 mm/10 m was used to
ensure a high density point cloud of the beam elements [23].
3.1 Building Information Model

One section of the roof was modelled and 11 beams were isolated represented by 40
million points. The total beam length is approximately 26 m. After filtering stray
points, the point cloud of the section was unified with an average spacing of 3 mm,
resulting in 6 million points or 230 thousand points per meter. The Poisson meshing in
3D Reshaper was performed using the two step meshing algorithm explained in
Sect. 2. A sampling size of 5 mm was applied for the meshing. A total of 1.2 million
triangular surfaces were constructed, approximately 40 thousand triangles per meter.
The meshes were manually segmented. For computational efficiency, the mesh com-
plexity was reduced. A maximal deviation of 1 mm was allowed on the decimation,
resulting in 50 thousand triangles. An average deviation of 0.35 mm was computed for
the entire data set.
The meshes were reverse engineered to solid models employing the ScanTo3D
function in SolidWorks. On average, the elements were created with a 0.02 mm
deviation from the mesh model. The parametric objects were exported to IFC 2 × 3
without any loss of data. Within this model, additional non-metric properties can be
defined by the user.
For the wire-frame model, hand measurements were acquired at the best suited
location for the accessible elements. For the inaccessible elements, total station mea-
surements were used. Rectangular beams were employed to best approximate the
structural elements. Also, the elements are linear and are located in the same plane.
3.2 Structural FEM-Based Model

A structural analysis model is defined for the evaluation. The varying parameters such
as material properties will be determined based on the European Standards. Both
geometric models will be evaluated with the same parameters.
Material. Code-based values are assumed as a first approximation since a detailed

assessment of the timber’s strength properties, e.g. by means of non-destructive
techniques (NDT) and minor-destructive testing (MDT), is not the focus of this paper.
Therefore, rheological phenomena in timber are currently not taken into account. The
material properties are derived from the EN 338 standard for the strength categorisation
of structural timber [16]. The roof elements are macroscopically identified as hardwood
oak with an assumed strength class of D30 and an orthotropic behaviour (Table 1).
Table 1. D30 properties from EN 338 [16]

General Strength properties [N/mm2] Stiffness properties [kN/mm2] Density
[kg/m3]
Material Class fm,k ft,0,k ft,90,k fc,0,k fc,90,k fv,k Em,0,mean Em,0,k Em,90,mean Gmean qk qmean
Hardwood D30 30 18 0.6 24 5.3 3.9 11 9.2 0.73 0.69 530 640
oak
The material properties are affected by the duration of the load and the moisture
content. Factor kmod is introduced to deal with these effects. In our test design, the roof
structure is assumed to be service class 2, which corresponds to an average working
temperature of 20° and a relative humidity more than 85 % for only a few weeks per
year [18]. The matching kmod is 0.6 and 0.9 for the permanent and short-term load
durations respectively. The reduced characteristic strength value Xd is calculated using
Eq. (1). γM is the partial factor for the material property which is 1.3 for solid timber.
Xk
Xd ¼ kmod ð1Þ
cM
Joints. Varying types of connections are present in the roof structure: mortise and
tenon joints, and dovetail half lapped joints. Both are constructed with wooden dowels.
While these connections can resist normal forces, they have limited rotational stiffness.
Therefore, the connections are designed as perfect hinges. The joints that connect the
roof to the building are modelled as fixed joints. This decision is supported by field
observations. Also, the use of perfect hinges simplifies the calculations and overesti-
mates the bending moment and deflections, resulting in a conservative analysis [24].
Loads. The different loads imposed on the roof structure are the self-weight, the
imposed loads, the wind loads and the snow loads. The self-weight is considered the
dead load of the structure. The permanent load consists of the roof frame (rafters and
battens) and the tiles. The imposed load is neglected because of the inaccessibility of
the roof.
For the snow and wind loads, the angle of the roof is required. From the laser scan
data is derived that the roof has an angle of 58°. The Eurocode EN 1991-1-3 [19]
specifies that snow loads on such steep roofs can be neglected. For the wind load, only
the forces in the plane with the considered roof truss are considered. The wind pressure
is obtained using Eq. (2) [20]. qp is the peak velocity pressure, ze is the reference height
for the external pressure and Cpe is the coefficient for the external pressure. Category III
best fits the terrain category of the test site. The wind pressure on the strong side of the
section mounts up to 0.41 kN/m2. On the weak side a suction force of −0.51 kN/m2 is
calculated.
We ¼ qp ðze Þ Cpe ð2Þ
The roof is connected to the structural elements by purlins. Therefore, the varying
loads are represented by concentrated forces located perpendicular to the purlins at their
respective connections. Figure 5 shows the roofs connections in detail as well as the
location of the varying loads split into the vertical and horizontal components.
Fig. 5. Load model: the forces are transmitted from the rafters to the beams trough the purlins
(left). The wind load and the permanent load of the roof are depicted in red. The self weight of
the structural elements is shown in yellow (right). (Color figure online)
Structural Analysis. Both structural models are tested in the Serviceability Limit State
(SLS), which calculates the overall deformations of the structure [18]. It applies to
structures, components and connections with the assumption of a linear relationship
between the subjected actions and the corresponding deformations. The total defor-
mation is calculated for the worst combination of the loads.
4 Experimental Results
The comparison between the two models is performed in two stages. First a metric
comparison is made. Second, a structural analysis comparison is made.
Geometry. To compare both geometric models, a million sample points are generated
on the surface of the models. The shortest euclidean distance between the two
geometries is calculated in the CloudCompare software. The results are shown in
Fig. 6. A mean distance of 20 mm with a standard deviation of 40 mm is calculated
between both data sets. This results in a mean difference in profile area of 15 % and in
volume of 12.5 %. These numbers indicate a major discrepancy between both data sets.
The complex model has more weight because the modelling in the wire-frame model is
based on the smallest section. The mass discrepancy both affects the load from self
weight and the structural behaviour of the beams. Figure 6 shows that the largest
deviations are located in the more complex areas. These differences are the result of the
abstractions made by hand measurements. Moreover, the wire-frame model does not
encompass geometries such as complex connections, non-rectangular and varying
sections, non-linear beams and out of the plane elements. The difference in geometry is
expected to have a major impact on the structural analysis.
Fig. 6. Overview metric comparison: the smallest euclidean distance between both data sets at
each location (left), and the histogram of the errors (right).
Structural Analysis. Both models are tested with the same material parameters and
external loads specified in Sect. 3. Figures 7 and 8 depict the results from the SLS
analysis. They show the deformations of the elements for the varying loads on the
structure. The discrepancy in deformation between the models is very noticeable. The
complex model shows deformations up to three times larger than the maximum
deformation in the wire-frame model. This is partially caused by a difference in weight
and section profiles, as well as the difference in modelling between the two structural
models. As stated in the geometry comparison, the complex model consists of
non-rectangular beams that connect to the joints in varying angels at different locations.
This causes forces to be transferred sub-optimal. Also, the location of the loads has a
major impact. In the complex model, the loads are placed more accurately, causing a
difference in stress concentration.
In addition to an increased stress in the beams, both models behave differently. For
instance, the main deformation is located elsewhere in both models. In the wire-frame
model it is located on in the left upper beam, while in the complex model it is located in
the right upper beam. This is caused by the behaviour of the central horizontal beam.
From Fig. 7 is derived that in the wire-frame model, the horizontal beam in the middle
of the structure is a crucial component to transfer forces from the right to the upper left
side of the structure. In the complex model, this is not the case. If the deformations are
Fig. 7. Deformation wire-frame model. Largest deformation located in upper left beam.
Fig. 8. Deformation complex model. Largest deformation located in upper right beam.
compared to the deviations in the modelling (Fig. 6), it is revealed that this beam shows
the largest discrepancies. This is caused by a difference in modelling. In the wire-frame
model, this beam was modelled straight and fitted between the vertical beams. In the
complex model, this beam is out of the section plane in accordance with the real
conditions. Hence, the beam cannot transfer forces as efficiently in the complex model.
As a result, the peak deformation is located elsewhere in both models (Fig. 8).
5 Conclusion
In this paper we presented a method to create realistic BIM objects of heritage roof
elements and subsequently apply the geometric model as input for structural modelling.
Instead of utilising traditional wire-frame models, we used more complex solid models.
By employing accurate dense point clouds, we were able to construct as-built models of
the structural elements of a building. The test results proved that with our approach,
volumes are estimated 12.5 % more accurately and profile areas 15 % more accurately.
Furthermore, the complex solid model was able to encompass complex geometries
such as complex connections, non-rectangular and varying sections, non-linear beams
and out of the plane elements. The simulation results from the structural analysis
showed that these differences in geometry have a significant impact on the structural
behaviour of the timber roof truss. Deformations up to three times larger than the value
of the maximum deformation in the wire-frame model were measured in the complex
models. These observations are crucial in the preservation process. Additionally, our
method provides a solution to deal with data heterogeneity. By creating BIM objects,
we allow the exchange of data between different software and stakeholders in the
heritage process.
Acknowledgement. The research group Geomatics of the Department of Civil Engineering, TC

Construction and the support of the Raymond Lemaire International Center for Conservation, KU
Leuven are gratefully acknowledged.
References
1. Castagnetti, C., Bertacchini, E., Capra, A., Dubbini, M.: Terrestrial Laser Scanning for
Preserving Cultural Heritage: Analysis of Geometric Anomalies for Ancient Structures,
TS08C-Laser Scanners II, 6031 (2012)
2. Arias, P., Armesto, J., Lorenzo, H., et al.: Digital photogrammetry, GPR and finite elements
in heritage documentation: geometry and structural damages. In: ISPRS Commission V
Symposium ‘Image Engineering and Vision Metrology’ DIGITAL (2012)
3. Bassier, M., Yousefzadeh, M., Van Genechten, B.: Evaluation of data acquisition techniques
and workflows for scan to BIM. In: Proceedings of Geobussiness (2015)
4. Van Roy, N., Verstrynge, E., Brosens, K., Van Balen, K.: Quality management of structural
repair of traditional timber roof structures. In: Structural Analysis of Historical Construc-
tions, Anamnesis, Diagnosis, Therapy, Controls (2016)
5. Mngumi, E.A., Ruther, H.: Solid modeling for heritage documentation. In: Workshop –
Archaeological Surveys, WSA2 Modelling and Visualization (2004)
6. Castellazzi, G., Altri, A.M.D., Bitelli, G., Selvaggi, I., Lambertini, A.: From laser scanning
to finite element analysis of complex buildings by using a semi-automatic procedure.
Sensors 15, 18360–18380 (2015)
7. Son, H., Hwang, N., Kim, C., Kim, C.: Rapid and automated determination of rusted surface
areas of a steel bridge for robotic maintenance systems. Autom. Constr. 42, 13–24 (2014)
8. Saygi, G., Agugiaro, G., Hamamcolu-Turan, M., Remondino, F.: Evaluation of GIS and
BIM roles for the information management of historical buildings. ISPRS Ann.
Photogramm. Remote Sens. Spat. Inf. Sci. 2, 283–288 (2013)
9. Garagnani, S., Manferdini, M.: Parametric accuracy: building information modelingprocess
applied to the cultural heritage preservation. In: International Archives of the Photogram-
metry, Remote Sensing and Spatial Information Sciences, vol. XL-5/W1 (2013)
10. Logothetis, S., Delinasiou, A., Stylianidis, E.: Building information modelling for cultural
heritage: a review. ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci. II-5/W3, 177
(2015)
11. Murphy, M., McGovern, E., Pavia, S.: Historic building information modelling-adding
intelligence to laser and image based surveys of European classical architecture.
ISPRS J. Photogramm. Remote Sens. 76, 89–102 (2013)
12. Dore, C., Murphy, M., McCarthy, S., Brechin, F., Casidy, C., Dirix, E.: Structural
simulations and conservation analysis-historic building information model (HBIM). ISPRS
Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. XL-5/W4, 351–357 (2015)
13. Oreni, D., Brumana, R., Della Torre, S., Banfi, F., Previtali, M.: Survey turned into HBIM:
the restoration and the work involved concerning the Basilica di Collemaggio after the
earthquake (L’Aquila). ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci. 5, 267–273
(2014)
14. Barazzetti, L., Banfi, F., Brumana, R., Gusmeroli, G., Previtali, M., Schiantarelli, G.:
Cloud-to-BIM-to-FEM: structural simulation with accurate historic BIM from laser scans.
Simul. Model. Pract. Theor. 57, 71–87 (2015)
15. Kazhdan, M., Bolitho, M., Hoppe, H.: Poisson surface reconstruction. ACM Trans. Graph.
32(3), 1–13 (2013)
16. European Committee for Standardisation, EN 338: Structural timber-Strength classes (2016)
17. European Committee for Standardisation, EN 1991-1-1: Eurocode 1: Actions on structures -
Part 1-1: General actions - Densities, self-weight, imposed loads for buildings (2002)
18. European Committee for Standardisation, EN 1995-1-1: Eurocode 5: Design of timber
structures - Part 1-1: General - Common rules and rules for buildings (2014)
Part 1-3: General actions - Snow loads (2015)
Part 1-4: General actions - Wind actions (2005)
21. ANSYS, Inc.: ANSYS Modeling and Meshing Guide (2005)
22. Verstrynge, E., Kips, G.: Exploitation of 3D laser scanning data to generate FEM structural
models of historical timber roof structures (2016)
23. Bassier, M., Vergauwen, M., Van Genechten, B.: Standalone terrestrial laser scanning for
Effciently capturing AEC buildings for as-built BIM. ISPRS Ann. Photogramm. Remote
Sens. Spatial. Inf. Sci. III-6, 49–55 (2016). doi:10.5194/isprs-annals-III-6-49-2016
24. Branco, J., Varum, H.: Behaviour of Traditional Portuguese Timber Roof Structures (2006)
Implementation Analysis and Design for Energy Efficient
Intervention on Heritage Buildings
Elena Gigliarelli ✉ , Filippo Calcerano, and Luciano Cessari

( )
National Research Council CNR, Institute for Technologies Applied to Cultural Heritage ITABC,
Via Salaria km. 29.300, 00016 Monterotondo, Italy
{elena.gigliarelli,filippo.calcerano,
luciano.cessari}@itabc.cnr.it
Abstract. The study focuses on a multi-scale and multi-disciplinary approach,

for energy efficient intervention on the historic centre and buildings of a town in
southern Italy. The methodology involves the use of numerical simulations and
building information modeling for the management optimisation of the analysis
and design phases. The energy analyses are carried out with experimental meas‐
urements and numerical simulations and are integrated with traditional historical,
typological and architectural analyses. The study confirms the optimal behaviour
of historic settlement principle against new urbanisation and proposes a series of
specific solutions to be implemented on the buildings based on improving energy
efficiency and sustainability of interventions, compatibility with the restoration
charts and with the historical and microclimate context of reference. The study
highlighted a number of limitations still present in the interoperability between
software that future research developments will have to overcome in order to
improve the practical applicability of the approach.
Keywords: Cultural heritage · Building performance simulation · Numerical

simulation · BIM · Diagnostic · Urban historic fabric · Energy analysis
1 Introduction
1.1 Cultural Heritage and Sustainable Development
The energy consumption of buildings in Europe is about 40 % [1] of total energy demand
and about 36 % of total CO2 emissions [2] and a large part of this consumption (over
80 %) is due to the buildings during their life cycle and do not depends on their construc‐
tion [3]. In Europe more than 40 % of the building stock was built before 1960, in the
absence of laws on energy efficiency [2], and the average amount of investment in new
construction accounts for about 1 % of total investments in the construction sector [4].
There is therefore a need to act according to the principles of sustainability and energy
efficiency even on existing buildings [5] to make the historic fabric of European cities
less energy-intensive and more resilient to climate change [6, 7].

DOI: 10.1007/978-3-319-48496-9_8
92 E. Gigliarelli et al.
1.2 Conservation and Energy Revitalization Trend of Historic Architecture

Cultural heritage preservation and energy savings are two important issues for Europe
but at the moment there isn’t a strong coordinated action on the matter [8]. Energy
building retrofit is the application of the best set of technologies to achieve the greatest
reduction in energy consumption while maintaining satisfactory internal comfort condi‐
tions under certain operating constraints [5]. For proper energy efficient and sustainable
design of the built heritage conservation, in the decision making are added the issues of
the compatibility of the interventions and materials with the restoration charts and regu‐
lations in force and the evaluation of the intervention in its whole life cycle along with
its the socio-economic and environmental impacts [9]. The European Commission has
taken several actions to reduce energy consumption in buildings, through two directives
on the energy performance [1, 10], but the difficulties in finding energy efficiency meas‐
ures that do not compromise the architectural, historical and cultural values of the built
heritage has led lawmakers to exclude it from the laws on energy efficiency [11]. Never‐
theless minor and moderate interventions on historic buildings are growing and are
characterized by a significant potential for the reduction of energy consumption as
demonstrated in the European project SECHURBA [12]. The energy efficiency of the
cultural heritage is finally starting to get out of a “regulatory” dimension in contrast with
the professional practice and is becoming a strategy to protect the built heritage that in
any case is subordinated to its preservation [13, 14].
Generally the approach is characterized by two different scales: at the urban scale
are performed expeditious analyses on the fabric, while at the building scale are
performed detailed surveys and analyses and more accurate and targeted interventions
though they are often decontextualised from the microclimate reference condition [15].
For a proper energy efficient and sustainable design, the approach should instead be
simultaneous. The relationship between buildings and climatic and microclimatic
context is undeniable in historical architecture, both in construction materials and
methods and in the morphology of settlement and building typologies which are deter‐
mined by biophysical and bioclimatic conditions. Investigating and designing these
historical fabrics also from a bioclimatic point of view means not only to improve them
from an energy standpoint, but also to rebuild the natural functioning processes of
historical and architectural structures enhancing at the same time their distinctive char‐
acteristics and identities linked to the conditions of the local microclimate [16]. More‐
over, if the approach is also used for newly urban portions it allows for a better harmo‐
nization with the historical part.
1.3 Change to the Methods and Tools for Numerical Simulation and Assessment
An integrated and holistic multi-scalar approach requires new methodologies and

management tools. The research is then oriented towards the use of numerical simulation
(NS), and the integration of the various disciplinary contributions into a single IT plat‐
form.
Implementation Analysis and Design for Energy Efficient Intervention 93
The urban morphology has a substantial effect on the microclimate. Within the Urban
Canopy Layer (UCL) there is an interrelation between indoor and outdoor thermo-
hygrometric condition which reflects on the building energy balance and consumption
[17]. A quantitative study identifies energy savings achievable in buildings by opti‐
mizing the external microclimate between 20 and 30 % of the total [18]. Microclimate
numerical simulations allow to investigate the ante-operam conditions and to evaluate
and compare the performances of different design alternatives. It is still difficult to
directly transfer these results to specific Building Performance Simulation (BPS) that
are the creation of a “behavioural model” of a building, reduced to a certain level of
abstraction of the physical real world entities and connected phenomena [19], and this
issue is currently the subject of innovative research insights [20].
For the integrated management of traditional and simulative analyses, some research
experiences has started to use Building Information Modeling, that is an additional
instrument for the design, representation, production and management of the built envi‐
ronment [21], declined in the field of cultural heritage (HBIM).
The interpretation of HBIM proposed in this study points to an integration in the
BIM environment of the geometric complexity of a building or historic fabric (only
partly achievable with parametric object libraries derived from manuals and historical
data), along with all the multidisciplinary data resulting from the aforementioned design
approach of sustainable and energy efficient conservation interventions, which includes
the experimental data and energy modeling, configuring then a Historic Energy BIM
(HEBIM).
2 Methodology
2.1 The Advantages and Limits of Numerical Simulations and Historic BIM in
the Regeneration Process
Numerical simulations are now recognized as a key tool for the energy retrofit of the
building stock and the consequent reduction of greenhouse gas emissions because they
treat the buildings as a system of interrelated elements that can be optimized and not as
the sum of a number of separately designed and optimized sub-systems or components
[22]. The purpose of the simulations is to pursue the understanding of passive and active
behaviour of the buildings in its microclimatic context [23], thus accelerating the anal‐
ysis and increasing the efficiency of the design process, and enabling at the analytical
stage the understanding of a greater range of design variables, in the design phase the
quantitative verification of the effectiveness of the design alternatives through retroac‐
tive feedback [19]. The application of numerical simulations in historical contexts trig‐
gers a tremendous increase in complexity and at present, few studies focus on the
numerical simulation of historic buildings and their calibration [24]. The difficulties
mainly depend on the constructive and geometric specificity stratified over the years and
on the need to apply dynamic simulations, both for the study of the inertial behaviour
that usually historic buildings have, and for the study in the summer season where the
simplifications of the steady or quasi-steady-state currently used in energy efficiency
regulations are likely to produce misleading results.
The integration between numerical simulations and BIM, due to a lack of stand‐
ardization of the modeling process of numerical simulation [25] and interoperability
limits of current BIM platforms [26], it is still a complex process, that is not auto‐
mated and heavily dependent on manual steps [27, 28]. Despite the views of the
scientific community about the possible future integration of numerical simulation
and BIM are still discordant [26, 29], especially on the timing within which this
integration will be sufficiently automated, it is now clear that the research should
produce the greatest efforts in the direction of this integration [23]. Pursuing this
scope applying NS and BIM in historical contexts allows to test the limits and poten‐
tial of current approaches, offering the possibility of promoting their evolution
towards a more effective practical application.
2.2 NS and HBIM Integration Workflow (HEBIM)

From the analysis of the most recent applications of BIM to the cultural heritage, with
a majority of case studies in Italy [30, 31], emerges a common approach based on data
collection (generally through 3D laser scanner survey), then processed for the acquisi‐
tion into the BIM environment, to which then follows each time the creation of a custom
“mini-library” of parametric “historic objects” which are then related to the virtual
geometric model [32]. Then to the objects are added properties concerning the data
relating to historical, structural [33] or energy analysis [34], even if the energy analysis
is generally performed with simplified software integrated in the same BIM tool. For an
optimised Historic Energy BIM (HEBIM) workflow it is instead necessary to integrate
the best simulation software (typically external to the BIM tools), taking care not to
increase the parallel geometric modeling, caused by low interoperability of software,
while increasing at the same time the degree of model flexibility and feedback between
NS and BIM.
3 Application
3.1 Selection of a Sustainable Heritage Case. The Town of Frigento
This integrated approach made up of urban regeneration and energy and sustainable
redevelopment strategies and methodologies has been applied in the development of the
Project METRICS “Methodologies and Technologies for the Management and Requa‐
lification of Historic Centres and Buildings”. METRICS is an industrial research project
funded by the PON Research and Competitiveness 2007-2013, which aims to develop
methodologies and innovative and sustainable technologies in historic city centres, for
energy and environmental requalification of historic buildings and to study the resilience
of urban systems. The research was carried out by a team of scientific institutions and
companies led from the Campanian high technology research centre for sustainable
construction STRESS.
Frigento, a historic town in the province of Avellino characterized by a strong seismic
activity was chosen as a case study and a demonstrator project. In 1980 the earthquake
that swept across the territory of Irpinia caused massive damage to Frigento, still
noticeable through numerous voids inside the urban fabric. To repair the damage, several
actions were planned and implemented concerning demolition works, structural rein‐
forcement and even numerous reconstructions that have led to a transformation of some
parts of the urban fabric.
The old town is now depopulated and features a mix of original and reconstructed
buildings that mimic the historical forms with new materials and technologies. Some of
these works along with a new expansion have extended the urban fabric along the main
roads, according to settlement principles different than the traditional ones.
As a research guideline on the historic fabric, it was decided to operate both at the
urban scale, focusing the analysis on a homogeneous and most populated construction
sector, and at building scale by working on two of the most representative monumental
buildings of Frigento and two smaller buildings with less historic and architectural value.
Testa-Cipriano palace, an elegant building with courtyard currently inhabited by the
owners, and De Leo palace that today houses the archaeological museum, the municipal
library, a conference room and a multifunctional area, were chosen as examples of the
monumental buildings of Frigento.
As regards the buildings characterized by minor historic architectural value, a four-
block area in the heart of the town has been identified after several inspections and
following careful analysis of maps and documentation, supported by meetings with local
experts and technical staff of the municipality. Within the area the investigation was
deepened on two terraced residential buildings. The first building has withstood the
earthquake and has undergone seismic upgrading works, the other the second was the
subject of heavy reconstruction by replacement of the original structure which was in
masonry with a new reinforced concrete (Fig. 1).
Fig. 1. Frigento, vico San Giovanni, volumetric assumption previous the earthquake of 1980 on
the left and current situation on the right
3.2 From the Survey to the Models: Data Integration of Energy and Restoration
Documentation
For the project, a multi-scale approach was adopted to integrate all the collected infor‐
mation in a series of federated models in HEBIM environment.
On the urban scale, historical, architectural, typological and microclimate studies

have been carried out with a first simulation run concerning the sunshine and microcli‐
mate conditions (using computing fluid dynamic simulations for analysing the
microclimate) (Figs. 2 and 3).
Fig. 2. Solar radiation winter season
Fig. 3. Winter average wind speed simulation, typical design day: 2017.02.07, h: 14:00
On the identified area further investigations have been developed. As shown in the
diagram below, from a 3D laser scanner survey (Figs. 4 and 5), geometric modeling has
been developed for BIM (Fig. 6). Inside the BIM an optimized database (Fig. 7)
containing historical and urban planning information has been linked to the geometry
(Fig. 6). The geometry was then adapted (simplified) to the level of detail required by
energy simulation, eliminating elements of excessive detail arising from the point cloud
of the laser scanner survey, or elements of the BIM environment too burdensome for
the simulation (such as the trees of BIM libraries, replaced with simplified forms). The
geometry was then sent to the simulative analysis of the external microclimate, from
which came back the graphical output to be displayed in the BIM model (Fig. 8).
Fig. 4. Workflow scheme (geometric flows highlighted in green, data flows in orange), 3D laser
scanning of the area (STRESS) (Color figure online)
Fig. 5. Stages of exploratory and diagnostic investigations: 3D laser scanning survey (STRESS);
heat flow analysis (STRESSS – TRE); environmental monitoring (STRESS, ITABC and IRISS
CNR, ETT, TRE, TME)
Fig. 6. HBIM modeling with integration of the urban plan in force

Fig. 7. Urban plans and typological and historical analysis of the reference area
Fig. 8. Simulation output shown in HEBIM environment
The modeling was then deepened over the selected buildings by associating to the
geometry functional layers, heat flow and thermographic survey and detailed simulations
of solar radiation on the surfaces (Figs. 9 and 10).
Fig. 9. HBIM model of one of the selected historic buildings
Fig. 10. Thermographic survey and sunshine radiation simulation shown in HEBIM
At the same time a research for energy efficient and sustainable materials and tech‐
nologies was carried out. The individual elements of the database have been assigned
scores by a team of experts according to their potential for improving energy perform‐
ance of the building, resource consumption, affordability and compatibility with the
restoration charts. The survey was then used as a decision support for selecting the
retrofit solutions for the selected buildings in order to achieve and surpass the Italian
regulatory standards aiming at an environmental sustainability of the intervention
including life-cycle cost analysis and social and environmental impacts.
4 Findings and Discussion
4.1 Energy Performance at Historic Urban Level
The results of the microclimate simulations confirm that the old urban fabric is organised
for an optimized seasonal behaviour, sheltering the prevailing winds and improving solar
exposure in the winter season. On the contrary, the areas of new construction are exposed
to the prevailing winds and with poor sun exposure in the cold season (Fig. 11).
Fig. 11. Summary of environmental performance of the municipality of Frigento in the winter
and summer seasons.
4.2 Energy Performance of Heritage Buildings

In all investigated buildings, the analyses have shown poor energy performance due to
poor insulation of the envelopes and ground floors of the buildings, poor moisture
management and inefficient systems. The envelopes of the historic buildings are slightly
less isolated but can count on a better thermal mass that provides better performances
in summer season.
For one of the investigated case studies, a retrofit intervention has been proposed
with the following design solution:
• the improvement of the envelope and roof insulation through natural sustainable
materials that do not hinder the normal transpiration of the building;
• the realisation of a photovoltaic field on the roof made of local innovative photovol‐
taic and photocatalytic tiles produced with sustainable production processes;
• the substitution of the existing boiler with high efficiency condensing boiler and the
realisation of an insulated radiant floor system for the ground floor.
5 Conclusion
5.1 Evaluation
The study, with a multi-scale and multi-disciplinary approach, focused on the applica‐
tion of new tools both for investigation and assessment of the conservation design
management in a town in southern Italy. The urban and building scales have been inte‐
grated in a progressive deepening of the investigation always able to maintain a strong
link between the different levels of the research.
The energy analyses have been carried out both with experimental measurements
and with numerical simulations and have been integrated with the traditional historical-
typological and architectural analysis.
The data produced were then integrated into a HEBIM platform to facilitate and
optimize their management and consulting. The study made it possible to test an inno‐
vative workflow showing the potential of a simultaneous analysis and design method‐
ology and highlighting a number of pending limits concerning interoperability between
software.
5.2 Future Research

Future research developments will concern the use of dynamic building performance
simulations on case studies in order to test and increase interoperability among numerical
simulations and BIM both in terms of geometry modeling and in terms of data visuali‐
sation of the simulation output inside the BIM environment.
Acknowledgments. This research is funded by the PON Campania Research and

Competitiveness 2007–2013. We also thank our colleagues from the research team led from the
Campanian high technology research centre for sustainable construction STRESS.
References
1. European Parliament: Directive 2010/31/EU of the European Parliament and of the Council
of 19 May 2010 on the Energy Performance of Buildings (2010)
2. Economidou, M., Atanasiu, B., Despret, C., Maio, J., Nolte, I., Rapf, O.: Europe’s buildings
under the microscope. A country-by-country review of the energy performance of buildings.
Buildings Performance Institute Europe, BPIE (2011)
3. Menassa, C.C.: Evaluating sustainable retrofits in existing buildings under uncertainty. Energy
Build. 43, 3576–3583 (2011)
4. Baek, C.-H., Park, S.-H.: Changes in renovation policies in the era of sustainability. Energy
Build. 47, 485–496 (2012)
5. Ma, Z., Cooper, P., Daly, D., Ledo, L.: Existing building retrofits: methodology and state-of-
the-art. Energy Build. 55, 889–902 (2012)
6. IPCC: 2014: Summary for policymakers. In: Field, C.B., Barros, V.R., Dokken, D.J., Mach,
K.J., Mastrandrea, M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C.,
Girma, B., Kissel, E.S., Levy, A.N., MacCracken, S., Mastrandrea, P.R., White, L.L. (eds.)
Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral
Aspects. Contribution of Working Group II to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change (2014). http://www.ipcc.ch/pdf/assessment-
report/ar5/wg2/ar5_wgII_spm_en.pdf
7. Leissner, J., Kilian, R., Kotova, L., Jacob, D., Mikolajewicz, U., Broström, T., Ashley-Smith,
J., Schellen, H.L., Martens, M., van Schijndel, J., Antretter, F., Winkler, M., Bertolin, C.,
Camuffo, D., Simeunovic, G., Vyhlídal, T.: Climate for culture: assessing the impact of
climate change on the future indoor climate in historic buildings using simulations. Herit. Sci.
3, 38 (2015)
8. Mazzarella, L.: Energy retrofit of historic and existing buildings. The legislative and
regulatory point of view. Energy Build. 95, 23–31 (2015)
9. Moropoulou, A., Labropoulos, K.C., Delegou, E.T., Karoglou, M., Bakolas, A.: Non-
destructive techniques as a tool for the protection of built cultural heritage. Constr. Build.
Mater. 48, 1222–1239 (2013)
10. European Commission: Directive 2002/91/EC of the European Parliament and of the Council
of 16 December 2002 on the Energy Performance of Buildings (2003)
11. Martínez-Molina, A., Tort-Ausina, I., Cho, S., Vivancos, J.-L.: Energy efficiency and thermal
comfort in historic buildings: a review. Renew. Sustain. Energy Rev. 61, 70–85 (2016)
12. Cessari, L., Gigliarelli, E.: Heritage-led eco-regeneration: the case of Zhejiang water towns
protection, restoration and preservation. In: Ioannides, M., Fritsch, D., Leissner, J., Davies,
R., Remondino, F., Caffo, R. (eds.) Progress in Cultural Heritage Preservation, pp. 369–377.
Springer, Heidelberg (2012)
13. Carbonara, G.: Energy efficiency as a protection tool. Energy Build. 95, 9–12 (2015)
14. De Santoli, L.: Reprint of “guidelines on energy efficiency of cultural heritage”. Energy Build.
95, 2–8 (2015)
15. Davoli, P., Belpoliti, V., Boarin, P., Calzolari, M.: Innovative methods for a sustainable retrofit
of the existing building stock a cross-path from social housing to the listed heritage. In:
TECHNE 08 (2014)
16. Baiani, S.: I criteri per il progetto, Cultural Technology. E Progetto Sostenibile Idee E Propos,
Ecocompatibili (2010)
17. Bouyer, J., Inard, C., Musy, M.: Microclimatic coupling as a solution to improve building
energy simulation in an urban context. Energy Build. 43, 1549–1559 (2011)
18. de la Flor, F.S., Domı́nguez, S.A.: Modelling microclimate in urban environments and
assessing its influence on the performance of surrounding buildings. Energy Build. 36, 403–
413 (2004)
19. Augenbroe, G.: Trends in building simulation. Build. Environ. 37, 891–902 (2002)
20. Calcerano, F., Martinelli, L.: Numerical optimisation through dynamic simulation of the
position of trees around a stand-alone building to reduce cooling energy consumption. Energy
Build. 112, 234–243 (2016)
21. Murphy, M., McGovern, E., Pavia, S.: Historic building information modelling – adding
intelligence to laser and image based surveys of European classical architecture. ISPRS J.
Photogramm. Remote Sens. 76, 89–102 (2013)
22. Hensen, J.L.M.: Towards more effective use of building performance simulation in design.
In: Leeuwen, J.P., Van Timmermans, H.J.P. (eds.) Proceeding of 7th International Conference
Design and Decision Support Systems in Architecture and Urban Planning, 2–5 July 2004.
Eindhoven University of Technology, Eindhoven (2004)
23. Clarke, J.A., Hensen, J.L.M.: Integrated building performance simulation: progress, prospects
and requirements. Build. Environ. 91, 294–306 (2015)
24. Roberti, F., Oberegger, U.F., Gasparella, A.: Calibrating historic building energy models to
hourly indoor air and surface temperatures: methodology and case study. Energy Build. 108,
236–243 (2015)
25. Hitchcock, R.J., Wong, J.: Transforming IFC architectural view BIMs for energy simulation.
In: Proceedings of Building Simulation 2011: 12th Conference of International Building
Performance Simulation Association, 14–16 November 2011
26. Ivanova, I., Kiesel, K., Mahdavi, A.: BIM-generated data models for EnergyPlus: a
comparison of gbXML and IFC Formats. In: Proceedings of Building Simulation
Applications, BSA 2015 - 2nd IBPSA-Italy Conference, Bozen-Bolzano, 4th – 6th February
2015 (2015)
27. Maile, T., O’Donnell, J., Bazjanac, V., Rose, C.: BIM-Geometry modelling guidelines for
building energy performance simulation. In: Proceedings of BS2013: 13th Conference of
International Building Performance Simulation Association, Chambéry, France, 26–28
August 2013
28. O’Donnell, J., Maile, T., Rose, C., Marazovic, N., Morrissey, E., Regnier, C., Parrish, K.,
Bazjanac, V.: Transforming BIM to BEM: generation of building geometry for the NASA
Ames sustainability base BIM (2013)
29. Bazjanac, V.: IFC BIM-based methodology for semi-automated building energy performance
simulation. In: 25th International Conference on Information Technology in Construction
(2011)
30. Logothetis, S., Delinasiou, A., Stylianidis, E.: Building Information modelling for cultural
herigate: a review. In: 25th International CIPA Symposium 2015, ISPRS Annals of the
Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. II-5W3, Taipei,
Taiwan, 31 August – 04 September 2015 (2015)
31. Garagnani, S., Manferdini, A.M.: Parametric accuracy: building information modeling
process applied to the cultural heritage preservation. ISPRS Int. Arch. Photogramm. Remote
Sens. Spat. Inf. Sci. 5, 87–92 (2013)
32. Dore, C., Murphy, M., McCarthy, S., Brechin, F., Casidy, C., Dirix, E.: Structural simulations
and conservation analysis -historic building information model (HBIM). In: International
Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol.
XL-5W4 2015, 3D Virtual Reconstruction and Visualization of Complex Architectures, Avila
Spain, 25–27 February 2015 (2015)
33. Oreni, D., Brumana, R., Della Torre, S., Banfi, F., Barazzetti, L., Previtali, M.: Survey turned
into HBIM: the restoration and the work involved concerning the Basilica di Collemaggio
after the earthquake (L’Aquila). In: ISPRS Annals of the Photogrammetry, Remote Sensing
and Spatial Information Sciences, vol. II-5, 2014, ISPRS Technical Commission V
Symposium, 23 – 25 June 2014, Riva del Garda, Italy (2014)
34. Brumana, R., Oreni, D., Raimondi, A., Georgopoulos, A., Bregianni, A.: From survey to
HBIM for documentation, dissemination and management of built heritage: the case study of
St. Maria in Scaria d’Intelvi. In: Digital Heritage International Congress (Digital Heritage),
pp. 497–504 (2013)
Historic BIM in the Cloud
Luigi Barazzetti(&), Fabrizio Banfi, and Raffaella Brumana
Department of Architecture, Built Environment and Construction

Engineering (ABC), Politecnico di Milano, Via Ponzio 31, 20133 Milan, Italy
{luigi.barazzetti,fabrizio.banfi,
raffaella.brumana}@polimi.it
Abstract. In this paper, we present a procedure which makes available an

accurate historic BIM (HBIM) in the cloud. Data processing is carried out with a
NURBS-based strategy to reduce the size of the final HBIM derived from
images and laser scans, providing an accurate and reliable 3D model with
limited memory occupation. This guarantees a remote access with PCs and
mobile devices connected through a cloud service.
Keywords: Accuracy AEC industry BIM Cloud Interoperability

NURBS
1 Introduction
Building Information Modeling (BIM) is becoming more and more important in the
Architecture, Engineering and Construction industry [1, 2, 7]. CAD-based drawings
(such as plans, sections, elevations generated by architects, engineers, designers, …)
are replaced by a three dimensional model of the building, from which standard 2D
project boards are automatically extracted. On the other hand, BIM is much more than
a 3D model. BIM requires parametric models where objects have relationships to other
objects (e.g. a door fits into a wall) and attributes (e.g. thermal properties, materials,
etc.) [9, 13].
Historic Building Information Modeling (HBIM) is a novel approach for docu-
mentation, conservation and assessment of historic buildings. It refers to the use of
BIM technology for modeling historic structures surveyed with laser scans and pho-
togrammetric methods [10, 14]. The aim of this work is to illustrate a procedure for
HBIM generation and the following use with a cloud service (Fig. 1).
The complexity of historic structures makes (parametric) modeling very difficult for
the lack of algorithms and procedures able to accurately reconstruct 3D shapes. Gen-
erating and managing an accurate HBIM can become a challenge when complex
elements cannot be simplified with basic geometric shapes [3]. In fact, most 3D
reconstruction approaches tend to generate meshes from point clouds to create a 3D
model from photogrammetric and laser scanning data. On the other hand, 3D recon-
structions based only on meshes are not optimal for BIM because of problems with
mesh parametrization [6, 8, 11, 12, 15, 18, 20, 21] and memory occupation.
A mesh is mainly defined as a set of connected 3D points. It is a static concept that
does not satisfy all the requirements of BIM projects. In addition, huge point clouds can
be needed in large construction projects, from which huge meshes can be generated.
DOI: 10.1007/978-3-319-48496-9_9
Historic BIM in the Cloud 105
Fig. 1. HBIM generated from point clouds is managed by project teams with centralized access
to the project data in the cloud.
The proposed approach focuses on NURBS rather than meshes. NURBS are
mathematical functions defined in parametric form and can be interactively manipu-
lated to reconstruct complex shapes. The goal of this work is the use of models based
on NURBS curves and surfaces for HBIM to provide an efficient solution that takes
into consideration the basic requirements of the surveying project (metric scale,
accuracy, level of detail, etc.). At the same time, it provides an efficient solution for the
use of the model in interactive 3D modeling environments.
Finally, the proposed approach allows one to exploit cloud computing solutions.
The reduced size of the model based on a NURBS objects allows a fluid visualization
when users interact with the model, so that architects, engineers, conservators, arche-
ologist, etc., can collaborate in a single, central workspace.
106 L. Barazzetti et al.
2 The Proposed Procedure for Historic BIM
The use of phase-shift laser scanning technology for digital surveying allows the
acquisition of a huge amount of 3D points (billion points for large and complex
projects) with a precision better than 5 mm. This is sufficient for typical project boards
used in the AEC industry, such as plans, sections, and elevations. The translation of the
point cloud into a BIM is carried out considering the expected accuracy at a given
metric scale (m). This evaluation is carried out with the simple relationship
E = 0.2m (from cartography) that results in 1 cm at the scale 1:50, and 2 cm at 1:100.
These are the most common metric scales used in construction projects.
Figure 2 shows the span of a mediaeval bridge, in which it is clear the irregular
geometry of the construction. Direct modeling from point clouds allows one to generate
Fig. 2. The case study used in this paper is a span of a medieval bridge. The figure shows an
orthophoto of the front (top), a plan (bottom-left) and a section (bottom-right).
a reconstruction based on curves that follows laser scanning data, but BIM software
have not libraries able to provide accurate reconstruction of the unique geometry of the
bridge.
For this reason, the procedure presented by [5], that exploit NURBS curves and
surfaces [16], is used to create 3D parametric objects of the different structural elements
of the bridge. Figure 3 shows the NURBS-based modeling results for the intrados of
the vault compared to the laser scanning point cloud. The overall discrepancy in terms
of RMS is about 5 mm, that is more than sufficient for the metric scale 1:50. In this
case, the vault cannot be approximated with simple elements available in BIM pack-
ages (e.g. arches) because of the operational requirements of the project.
Fig. 3. Comparison between the point cloud and the NURBS surface generated with the
proposed procedure. The overall discrepancy is 5 mm in terms of RMS.
The NURBS surface was then parametrized by means of an offset [17] of the
intrados, so that the original NURBS surface is preserved and a dynamic object-based
representation of the complete object (including semantics and attributes) is achieved
(Fig. 4). This last step, i.e. the creation of the final BIM objects, is carried out by
exploiting both geometrical and functional requirements, in which the geometrical
problem is solved through a multilayer structure with multiple offsets of the intrados.
The reconstruction of the whole span was carried out with the proposed NURBS
based procedure with the identification of the different constructive elements. This is a
fundamental step for the creation of accurate object-based reconstructions that will be
used by different specialists.
Fig. 4. The dynamic BIM representation of the vault. The intrados surveyed through laser scans
has a new parametric representation in which the thinness is interactively modified in the
software database.
The final BIM model has a size of 17 MB when imported and saved in Autodesk
Revit. The choice of Autodesk Revit as final software is motivated by the need of a
single platform for all the specialists involved in the project (Fig. 5).
Fig. 5. The final HBIM of the span is made up of several BIM objects for the different
constructive elements.
Fig. 6. HBIM simultaneously available in multiple devices thanks to A360 cloud technology
(top). Object properties are still available so that the database is (partially) preserved.
Fig. 7. Technical drawings available for the users involved in the project become the marker for
triggering AR apps.
The reader is referred to [19] for more details about the presented case study.
Additional images illustrating the work carried out for the whole bridge are shown in
Fig. 9, at the end of the paper.
3 Accurate HBIM in the Cloud
Because basic requirements of BIM projects require a collaborative environment for

architects, engineers, customers, producers, etc., cloud-BIM is a very promising
technology to simultaneously connect people avoiding the release of multiple project
versions with possible inconsistencies. A centralized version of the model can be
remotely accessed by multiple operators who can view, inspect, and edit project files
without expensive hardware and software. Real-time communication can be carried out
between multiple specialists through chat and email notifications. Such solution is
Fig. 8. AR allows one to see hidden elements, such as foundations, riverbed, and a digital
reconstruction of the possible wooden structure used to build the vault.
already available in some cloud-based platforms, allowing multiple specialists to work

inside a highly collaborative environment.
On the other hand, cloud-BIM is still a relatively new technology with a limited use
in practical applications. The first scientific papers on cloud-BIM were published in
2011. The number of available systems has rapidly increased, leading to the actual
avaibility of several platforms such as Autodesk 360, BIM9, BIMServer, BIMx, etc. As
things stand at the present, cloud-BIM is mainly used for the design and construction
phase of the building, with limited application to maintenance, building lifecycle,
facility management and demolition.
In this work Autodesk® 360 (A360) was used to use the bridge project in the cloud.
Only a span of the bridge was considered and future work will focus on the whole
model. Autodesk 360 is defined as “a cloud-based platform that gives you access to
storage, a collaboration workspace, and cloud services to help you dramatically
improve the way you design, visualize, simulate, and share your work with others
anytime, anywhere”. The system provides 3 GB of free storage space, whereas a
monthly, quarterly or annual subscription includes 25 GB and the access to cloud
computing services for rendering, optimization, energy analysis and structural analysis.
The file size limit is 2 GB per file.
The supported web browsers are Google Chrome, Mozilla Firefox, Microsoft
Internet Explorer for 2D visualization (plans, sections, elevations, …), whereas 3D
Fig. 9. The BIM project of the whole bridge, more details about the reconstruction process are
illustrated in [19]. The work presented in this paper was carried out with a single span.
viewing is only supported by Google Chrome and Mozilla Firefox (Fig. 6 – top).
A mobile application was also developed to bridge the gap between the office and field.
The mobile app supports more than 100 file formats used in the AEC industry
including 2D and 3D (e.g. dwg, dwf, ipt, iam, idw, rvt, sldprt, sldasm, asm, nwd, nwc,
catpart, catproduct, f3d, etc.). It can access email attachments and files from Dropbox,
Box, Google Drive, OneDrive, iCloud, Buzzsaw, etc. Basic functions are available for
reviewing and navigating models through intuitive touch-based navigation (zoom, pan,
rotate) as well as tools for annotation, markup, and comment. It should be mentioned
that operations are not carried out by simply visualizing the geometry of a 3D model.
Objects properties (level, type, category, …) are provided by selecting the different
objects, so that the project database remains available.
4 Conclusions and Outlooks
HBIM in the cloud is probably one of the future developments of BIM technology. The
avaibility of the model in mobile devices, as well as functions and tools for navigating
and editing, is potentially able to bridge the gap between office and construction site.
As things stand at the present, only PCs with reasonable screen size can be used for
effective work in which the model and its database are handled. On the other hand,
mobile applications can have a primary role for in-situ productive work, where digital
documentation is today substituted by printed project boards. The development of
cloud-based solutions and the use of tablets will integrate (not substitute) the traditional
use of paper. Cloud technology requires also a particular attention on security issues.
The presented A360 application is provided through a protected infrastructure for cloud
computing, in which both individuals and companies can securely access data stored or
data in transit.
As data are immediately available in 3D, cloud BIM opens new opportunities also
for augmented and virtual reality (AR and VR). IN particular, AR allows a direct
connection of projects boards with the 3D model available in mobile devices. Figure 7
shows the plan of the span with the 3D model, both inspected inside the office.
Marker-less AR represents a real challenge with formidable expected results during
on-site exploitation. Parts that are not visible (foundations and river bed, i.e. covered by
water), as well as elements that are no longer available (the wooden structure used
during the construction of the vault), can be reviewed (Fig. 8). The potential for generic
construction sites relies on the visualization of hidden elements (e.g. MEP) or the
progressive construction stages of the construction, among the others.
A significant difference between the BIM available in the cloud and that used in the
AR examples is relating to information loss. The model for AR is no longer a BIM. It is
turned into a static model without attributes, i.e. it requires a preliminary conversion in
common formats for direct 3D modeling. The avaibility of a complete BIM also in AR
apps is a possible future innovation to make BIM technology directly available.
References
1. Arayici, Y.: Towards building information modelling for existing structures. Struct. Surv.
26, 210–222 (2008)
2. Azhar, S., Khalfan, M.: MAQSOOD: building information modeling (BIM): now and
beyond. Australas. J. Constr. Econ. Build. 12(4), 15–28 (2012)
3. Baik, A., Alitany, A., Boehm, J., Robson, S.: Jeddah historical building information
modelling “JHBIM” – object library. ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci.
2(5), 41–47 (2014)
4. Barazzetti, L.: Parametric as-built model generation of complex shapes from point clouds.
Adv. Eng. Inform. 30, 298–311 (2016)
5. Barazzetti, L., Banfi, F., Brumana, R., Previtali, M.: Creation of parametric BIM objects
from point clouds using NURBS. Photogramm. Rec. 30(152), 339–362 (2015)
6. Barazzetti, L., Banfi, F., Brumana, R., Gusmeroli, G., Oreni, D., Previtali, M., Roncoroni, F.,
Schiantarelli G.: BIM from laser clouds and finite element analysis: combining structural
analysis and geometric complexity. In: International Archives of Photogrammetry, Remote
Sensing and Spatial Information Sciences, vol. 40(5/W4), pp. 345–350 (2015)
7. Bryde, D., Broquetas, M., Volm, J.M.: The project benefits of building information
modelling (BIM). Int. J. Project Manage. 31(7), 971–980 (2013)
8. Brumana, R., Oreni, D., Raimondi, A., Georgopoulos, A., Breggiani, A.: From survey to
HBIM for documentation, dissemination and management of built heritage: the case study of
St. Maria in Scaria d’Intelvi. In: Digital Heritage International Congress, Marseille, France,
pp. 497–504 (2013)
9. Eastman, C., Teicholz, P., Sacks, R., Liston, K.: BIM Handbook: A Guide to Building
Information Modeling for Owners, Managers, Designers, Engineers and Contractors, 2nd
edn. Wiley, Hoboken (2011). 626 pages
10. Fai, S., Graham, K., Duckworth, T., Wood, N., Attar, R.: Building information modeling and
heritage documentation. In: 23rd International CIPA Symposium (2011). 8 pages
11. Fai, S., Filippi, M., Paliaga, S.: Parametric modelling (BIM) for the documentation of
vernacular construction methods: a BIM model for the Commissariat Building, Ottawa,
Canada. In: ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information
Sciences, vol. 2(5/W1), pp. 115–120 (2013)
12. Fai, S., Rafeiro, J.: Establishing an appropriate level of detail (LoD) for a building
information model (BIM) – West Block, Parliament Hill, Ottawa, Canada. ISPRS Ann.
Photogramm. Remote Sens. Spat. Inf. Sci. 2(5), 123–130 (2014)
13. Kensek, K., Noble, D.: Building Information Modeling: BIM in Current and Future Practice.
Wiley, Hoboken (2014). 432 pages
14. Murphy, M., Mccgovern, E., Pavia, S.: Historic building information modelling (HBIM).
Struct. Surv. 27(4), 311–327 (2009)
15. Oreni, D., Brumana, R., Cuca, B.: Towards a methodology for 3D content models. the
reconstruction of ancient vaults for maintenance and structural behaviour in the logic of BIM
management. In: 18th International Conference on Virtual Systems and Multimedia - Virtual
Systems in the Information Society, Milan, Italy, pp. 475–482 (2012)
16. Piegl, L., Tiller, W.: The NURBS Book. Monographs in Visual Communication. Springer,
Heidelberg (2012). 646 pages. ISBN 978–3-642-97385-7
17. Piegl, L., Tiller, W.: Computing offsets of NURBS curves and surfaces. Comput. Aided Des.
31, 147–156 (1999)
18. Thomson, C., Boehm, J.: Automatic geometry generation from point clouds for BIM.
Remote Sens. 7(9), 11753–11775 (2015)
19. Barazzetti, L., Banfi, F., Brumana, R., Previtali, M., Roncoroni, F.: BIM from laser scans…
not just for buildings: NURBS-based parametric modeling of a medieval bridge. ISPRS Ann.
Photogramm. Remote Sens. Spat. Inf. Sci. III-5, 51–56 (2016). doi:10.5194/isprs-annals-III-
5-51-2016
20. Georgopoulos, A., Oreni, D., Brumana, R., Cuca, B.: HBIM for conservation and
management of built heritage: towards a library of vaults and wooden beam floors. In:
Proceeding of XXIV International CIPA Symposium (2013)
21. Oreni, D.: From 3D content models to HBIM for conservation and management of built
heritage. In: Murgante, B., Misra, S., Carlini, M., Torre, C.M., Nguyen, H.-Q., Taniar, D.,
Apduhan, B.O., Gervasi, O. (eds.) ICCSA 2013. LNCS, vol. 7974, pp. 344–357. Springer,
Heidelberg (2013). doi:10.1007/978-3-642-39649-6_25
Building Information Modelling – A Novel Parametric
Modeling Approach Based on 3D Surveys of Historic
Architecture
Fabrizio Banfi ✉
( )
Department of Architecture, Built Environment and Construction Engineering (ABC),

Politecnico di Milano, Via Ponzio 31, 20133 Milan, Italy
[email protected]
Abstract. Building Information Modelling (BIM) appears to be the best answer

to simplify the traditional process of design, construction, management and main‐
tenance. On the other hand, the intricate reality of the built heritage and the
growing need to represent the actual geometry using 3D models collide with the
new paradigms of complexity and accuracy, opening a novel operative perspec‐
tive for restoration and conservation. The management of complexity through
BIM requires a new management approach focused on the development of
improve the environmental impact cost, reduction and increase in productivity
and efficiency the Architecture, Engineering and Construction (AEC) Industry.
This structure is quantifiable in morphological and typical terms by establishing
levels of development and detail (LoDs) and changes of direction (ReversLoDs)
to support the different stages of life cycle (LCM). Starting from different expe‐
riences in the field of HBIM, this research work proposes a dynamic parametric
modeling approach that involves the use of laser scanning, photogrammetric data
and advanced modelling for HBIM.
Keywords: BIM · Complexity · LoD · NURBS · Point clouds
1 Introduction
The innovation and development of Information Communication Technology (ICT) in

support of the Architecture, Engineering industry, Construction and Operations (AECO)
is characterized by the growing use of BIM [9, 12].
The digitized management of historic buildings, infrastructure and complex systems
bases its foundation on theoretical and operational processes in continuous development.
In this process, the level of transmissibility of knowledge is fragmented and not fully
assimilated by the professionals involved. Their software and support tools are not able
to accommodate the correct level of complexity causing the generation of simplified 3D
models (not corresponding to the as-built) or complex BIM for a subsequent manage‐
ment.
Research and introduction of new paradigms as complexity for the built heritage is
a basic need to develop operative aspects to support identification, assimilation and

DOI: 10.1007/978-3-319-48496-9_10
Building Information Modelling – A Novel Parametric Modeling 117
transmission of useful and valuable information for the preservation of cultural heritage
[2, 7]. Each application requires geometric and technical characteristics defined by
advanced modeling procedures that could lead to data loss if not property tuned. Nowa‐
days, modeling requires long generative practices and specific computer skills that
correctly interpret the complexity of the buildings, and then transmit them within an
operating process that consists of different interdisciplinary teams [3, 5].
The complexity is a fundamental prerogative for high transmissibility of the infor‐
mation [10]. It results essential to enhance the vast number of technical analyses for all
types of buildings [1, 14]. It can become a definable standard parameter addressed to
European and global directives for the next years in favor of BIM integration in design
and construction processes.
The objectives of this research are:
• Integrate 3D survey techniques and advanced parametric modelling, providing a real
starting point for the creation of as-built BIMs with different LoDs.
• Argue the usefulness of a methodological advanced modelling approach that auto‐
mates the generative process of HBIM favouring the control, management and trans‐
missibility of the information collected during the building’s life cycle.
• Show how BIM can support the process of designing, building, restructuring, main‐
tenance and analyses, through three case studies aimed at different ReversLoDs
approach for historic buildings.
• To discuss, analyse, identify gaps in the work carried out, and propose possible future
research lines.
The process is based on research work carried out in recent years in the field of HBIMs
of great historical value buildings in the Italian territory. It was supported by public
institutions, multinational companies and research groups that have made possible a
sensitive scientific growth in the protection of architectural heritage. The Italian case
studies which will be described are the Basilica of Collemaggio in L’Aquila, Castel
Masegra in Sondrio and finally the Azzone Visconti Bridge in Lecco. The chronologic
order imposed in this paper follows the creation of models, from oldest to most current
BIM, in order to highlight the improvement of the interchangeability with different LoDs
obtained from 2013 to 2016.
2 A Novel Parametric Modeling Approach
This research proposes a dynamic methodological approach based on the management

of complex 3D shapes managed and manipulated with a new initial LoDs obtained from
the 3D survey and with bidirectional levels of complexity (ReversLods) which provides
the integration of NURBS modeling and parametric applications to support life cycle
management (LCM). HBIM increased the level of knowledge of the building through
the integrated use of various applications, which can be mainly divided into two macro
families: pure modeling software like MC Neel® Rhinoceros, and Autodesk® Autocad
3DS MAX, which are able to fully investigate the morphological appearance of the
building, and BIM Platforms such as Autodesk® Revit and ArchiCAD Graphisoft®,
118 F. Banfi
where the typological differentiation and the ability to associate additional information
is the basic logic. The first family is characterized by a free modeling approach that
generates 3D elements and surfaces able to reproduce complex forms. The user can
create surfaces and solids starting from the basic generative elements like lines, NURBS
(Non Uniform Basis Splines) [11], and mesh [16]. The disadvantage is the absence of
tools that allow the parameterization of the elements. The second family is characterized
instead by an internal structure made up of a database with families of objects. These
objects represent the architectural elements that form a building. Each element is adjust‐
able in its sub components, the parameterization of the dimensional components can be
adjusted and the type of each element can be edited through the use of settable parameters
(before and after) [6]. The main disadvantage that impedes the use of BIM for historic
buildings is the modeling of complex shapes: only through hard, long and complex
modeling practices it is possible to reach a sufficient level of detail. In historic buildings
morphological and typological aspects are not aligned. This is a fundamental problem
for the creation of HBIMs derived from 3D surveying techniques. The interoperability
of modeling software is a crucial aspect in order to optimize the peculiarities of the two
modeling families.
The transition from pure modeling software like Rhinoceros, AutoCAD to Revit
depended on the determination of the complexity and the level of accuracy of BIM.
Modeling must adapted to the information collected during the survey, allowing the
association of data to 3D parametric objects [4]. The advantage of combining these two
macro families with a survey based laser scanner (integration) could develop new meth‐
odological approaches to investigate the accuracy of existing assets and the overall
control over the entire design process as well as construction, maintenance of building
(Fig. 1).
Fig. 1. Generation process characterized by the integration of pure modeling (NURBS) with the
parametric modeling in BIM application.
For this reason, the 3D model must guarantee in terms of morphological features and
reality and then know how to steer a specific LoD [8]. The peculiarities of an existing
building, complex shapes of architectural and structural elements can be properly inter‐
preted, generated and represented by generative advanced modeling tools based on
NURBS technology.
The technological advancement requires continuous adaptation of the various stand‐
ards. In recent years, new parameters have been defined to assess the level of the 3D
model in geometrical terms and its associated information.
The management of CAD formats presupposes a development of interoperability

techniques to convey the totality of information available. This integration requires an
orientation of the models based on different LoDs of BIM, guaranteeing the new hier‐
archy of the model. This means that the use of the information associated with the para‐
metric model determines level of transmissibility of information of the building. Conse‐
quently, transmission of knowledge is determined by the modeling of each element and
by the level of complexity transmitted during the process. Figure 2 shows the integration
process used to generate HBIM.
Fig. 2. The proposed parametric modeling approach – integration of pure and parametric
modeling in order to improve the transmission of complexity, information and knowledge of the
building
3 LoD Level of Development and Detail
LoDs allow us to measure the reliability and security of information associated to the
BIM during the building process: starting from planning and construction to mainte‐
nance [8]. From an operational point of view, the advantages in the built environment
can be substantial especially in terms of time in the implementation phase of the 3D
model. LoDs are divided into a first ‘not modeled’ Conceptual representation (100), a
second level with the general requirements and the main quantities (200), a third level
to support the detailed design, a fourth level ‘manufacture/install’ in support of construc‐
tion planning (400), and a fifth and final level ‘as-built’ in support of a Asset management
(500).
In most cases, a 300-definition level can be a good initial compromise to manage a
new building. It is possible to pass to the next levels following the requirements of the
design and the following integration in the building process. It is sometimes impossible
to define a correct LoD for existing and historic buildings. For example: the use of a
laser scanner does not allow the automatic recognition of materials and elements.
120 F. Banfi
On the other hand, the protection and conservation of the built heritage requires,
(during the survey) to use non-invasive methodologies. The use of destructive techni‐
ques, such as core drilling, is not permitted in order to protect the building. Therefore,
the materials used and the construction techniques are unknown in the early stages of
laser scanner survey. The survey of the building can reveal morphological and typo‐
logical aspects even after years. However it is sometimes necessary to decrease the level
of definition of the model obtained, reversing the accuracy of the model to support the
design process. Consequently, it is not possible to define a clear LoD during the initial
agreement between BIM developer and customer; therefore the model will consist of
elements with different LoDs. This should require a further phase of definition of not-
modeled objects. The transmission clearly implies a duty to report the deficiencies of
the model in terms of quality and quantity.
The following case studies show the need to integrate different LoD in the generation
phase of the models to support the process and the importance of advanced modeling
that requires to invert the level of modelling (ReverseLoDs) for specific parts/compo‐
nents of the building.
4 BIM Process: Development and Management
The proposed method is composed at different operative steps that required different
types of model. Modelling has to guarantee different sub-models starting from the same
model with different levels of precision, accuracy, definition and development (Fig. 3).
Fig. 3. Orientation model based on the integration of advanced modeling to improve the
transferability of knowledge in BIM
The generation of complex elements such as vaults, arches, domes and walls
damaged capable of represent the laser data requires complex modeling practices.
Figure 4 describes the process in its analytic and operative phases. It shows the
required model and the ReverseLoDs approach:
Fig. 4. BIM Implementation: different type of model in life cycle
The key idea is based on both interoperability and integrated use of development
levels, as well as definition (for each building) should be targeted of specific analyses.
The analysis software must be viewed from the point of view of an advanced modelling
practice, that is a constraint in which the first modeling step must be oriented to the
correct reading of the BIM. It must to update and change throughout the process and
this requires an improvement of knowledge of the modeling techniques. This requires
a model that must allow an increase or decrease (ReverseLoD) in the level of complexity.
The new paradigm of Orientation of model/modelling is crucial to achieve such partic‐
ular type of analyses [15].
122 F. Banfi
5 Case Study 1: Bidirectional Generative Process to Support the

Restoration Project of Basilica of Collemaggio in L’Aquila
In 2009 the Aquila’s earthquake severely damaged the apse of the basilica, leading to
the collapse of the roof. The pillar of the transept completely collapsed. Their supported
the weight of the vault and the covering system. Naves have suffered subsidence in the
north wall (out of plumb of about 40 cm) and in the pillars. This case study shows that
the levels used in the model are different because they had to support different phases
of the restoration project. Specifically, the generative phase resulted in an HBIM and its
use in the restoration project, with a special attention to procedures used to preserve the
complexity of photogrammetry and laser scanning data and not always suit the design
phase. The 3D survey of the basilica required integration of detection techniques
(Fig. 5) that could restore the morphological complexity of the construction.
Fig. 5. 3D survey campaign with UAV, total station and laser scanner 3D
The geodetic network was measured with a Leica® TS30 total station, with a level
of precision on the measurement of 0.6 mm and 0.15 mgon on measures of azimuth and
zenith angles. The number of scans was 182 collected, with a resolution of 44 million
points. The UAV survey with Astec Falcon 8 equipped with RGB camera Sony NEX -
5N photogrammetrically calibrated performed have allowed us to use a powerful tool
to inspect the condition of the roof in 2013, providing useful data for the generation of
as-built model of the basilica.
The morphological complexity of the Basilica was the key to refine modelling tech‐
niques oriented to the generation of a BIM that represent the actual reality of the building
adapted to project requirements. The main objective was to define a high level of detail
of the constructive elements of the basilica and achieve an accuracy of modeling with
parametric surfaces and objects with a LoD 500. The architectural elements generated
from point clouds were characterized obtaining semi-automatic generative profiles
through the integrated use of software McNeel® Rhinoceros and plug-in Pointools. This
is a support for managing and editing the clouds to proceed to the realization of 2D
drawings (plans, elevations and sections) and then the model. Figure 6 shows the gener‐
ation process: the slice extraction procedure from point clouds led automatic generation
of NURBS surfaces and solids. They were obtained by a preliminary morphological
quantification corresponding to a LoD 300.
Fig. 6. Generative process. From pointclouds to 3D model the pillars defined by each single
stone. The HBIM reaches LoD 500
The flexibility of the method, has allowed transferring each element generated with
the pure modeling at a LoD 400, except the excavations, cellars and external parts still
characterized by a LoD 200 for a lack of data on the environments, resulting in a lack
of thickness of the walls and only a 3D generation of intrados surfaces. Starting from
the use of Autodesk® Revit, the HBIM of the Basilica was able to associate useful
information to the next steps. Direct survey was essential to determine each sub stone
form. Each pillar is composed of a certain number of courses; each course is composed
of a variable number of stone blocks (Fig. 6). The definition of the laser scanner survey
did not allow detecting the joints of connections of the various segments because of the
safety straps covering each pillar, so the direct survey allowed migrating the main
geometric information of the blocks in the HBIM with LoD 500.
Subsequently, the level of definition achieved exceeded the real needs for the
Construction Site (4D) of this phase [13]. The simplification required a LoD
200/300/400.
Finally, the flexibility of the 3D model allowed the updating of the project
requests in the preliminary and final phase thanks to 3D rendering and digital video
simulations generated with the application Abvent® Artlantis. As shown in Fig. 7,
Fig. 7. Comparing as/built and project. Through different LoDs it was possible to integrate BIM
(LoD 500) with 3D design integration of the Basilica with a LoD 300.
124 F. Banfi
through ReverseLods it was possible to integrate design assumptions (LoD 300) to

LoD 500 (as-built) of the basilica, supporting the design choices during life cycle of
project restoration.
6 Case Study 2: Modelling, NURBS-Mesh Technology, LoDs and

HBIM of Masegra Castel in Sondrio
This case study involved the construction of a HBIM of Castel Masegra in Sondrio by
laser scanner and photogrammetric data. The objective of this project is to preserve the
architectural and landscape heritage of Valtellina and Valposchiavo, sharing activities,
knowledge and methodologies using the most advanced techniques for diagnostics,
restoration, maintenance and management of cultural heritage, which were developed
by interpreting the preservation as a long-term process. The 3D survey has produced
176 scans for a total of 7.5 Billion of points with a resolution greater than 3 mm by using
a Faro® Laser Scanner Focus 3D and Leica® TS30 total station. Each scan was
composed of 44 million points. A robust geodetic network controlled the large amount
of data collected with laser scanner.
The modeling phase is based on a generative approach of the various elements
surveyed with the laser scanner. The integrated use of Autodesk® AutoCAD and
McNeel® Rhinoceros allowed us to generate 3D profiles for the generation of NURBS
surfaces able to follow the clouds of points with an average deviation of around 2, 3 mm.
Specific tools for point cloud management were recently included in AutoCAD2015.
The integrated use of the Pointools plugin in McNeel® Rhinoceros and Autodesk®
ReCap have greatly shortened the generative process of the model. Cloud management
was carried out through the new .rcp format, which can index and group all the scans.
New cleaning tools of point clouds have determined the area of interest directly in
AutoCAD, without using other plugins. The case study of the umbrella vault at the
second floor exploited the entire method applied to the Basilica of Collemaggio (Fig. 8).
Fig. 8. Modeling process – integration, orientation, LoD and interoperability of 3D model
The combined use of laser scanning data and Agisoft® PhotoScan gave an ortho‐
photo that highlights the morphological complexity of the vault. It was generated by the
alignment of 34 photos and a total of 95110 points. The use of a total station allowed to
define 9 control points and correctly georeferenced the photogrammetric model with the
laser point clouds. The integration of these two techniques allowed us to reprocess data
collected directly in Rhinoceros and model the data by importing the mesh generated in
Agisoft® PhotoScan through the obj file format. Through this format all the information
can be listed for the definition of lines, polygons and curves and complex surfaces. 3D
mesh generation led to determine a number of profiles and control sections of the
complex geometry of the vault in McNeel® Rhinoceros. To avoid possible geometrical
deviations, each profile was used with the ReverseLoDs approach, for the next portion
of the vaults.
Thanks to the dynamic offset, it was possible to automatically parameterize surfaces
with a LoD 400. Figure 9 describes the modelling process used: the transformation of
wall surfaces allowed the integration of physical information, for materials with LoD
500 and the orientation the model for different analyses.
Fig. 9. From 3D survey to BIM LoD 500 and an example of historical rib
7 Case Study 3: LoD 500 for Defining the HBIM of a Mediaeval

Infrastructure - Visconti Bridge in Lecco, Italy
The third case study is the HBIM of the medieval bridge Azzone Visconti in Lecco. This
is still under investigation by the public administration of the town of Lecco. The main
objective of the research was to create a model that can accommodate the maximum
amount of information collected by other research groups (geologists, structural engi‐
neers, historians etc.) to monitor and preserve the bridge.
The survey campaign involved the generation of a geodetic network of 77 scans
obtained with a Faro Focus 3D. The total number of laser scanning points was more than
2.5 billions, with an average precision of ±3 mm achieved by using chessboard targets
detected with the total station and additional scan-to-scan correspondence (spherical
targets). Photogrammetry has allowed the realization of number orthophotos of the 11
arches of the bridge and its fronts.
The purpose of modeling was to achieve a LoD 500 and implement the method of
the two cases previous studies. The scans of the bridge have been transferred directly
into the Mc Neel® Rhinoceros. The NURBS technology (due to its ability to interpolate
points) allowed us, to generate surfaces that follow the data laser without geometric
deviations. The creation of the wall partitions in Autodesk® Revit carried out through
126 F. Banfi
the automatic parameterization of the NURBS surfaces using the tools Wall By Face
and Edit Profile. Thanks to the resolution of the point clouds has been possible to get an
example of historical rib with LoD500. The putlog hole have determined the thickness
of the various wooden elements of the rib. With the 3D Augmented app you can directly
simulate the insertion of the rib on the river bridge.
A comprehensive database (LoD 500) for objects and share information for portable
devices was created with Autodesk® A360. Finally, Abvent® Artlantis and iVisit 3D
allowed us to created a virtual simulation of the medieval system of construction. The
transmission of information to structural engineers, geologists and supervisors can take
place by importing the whole model in a cloud service with the automatic ReverseLoDs
approach at the level 500.
8 Results and Discussion
The HBIM the Basilica of Collemaggio can directly demonstrate the general complexity
of geometric model through different LoDs, which avoids simplified representations that
do not match the actual complexity of the building. In the Concept and Design phase
the model can receive changes in certain areas of the building to support different design
solutions. The inversion of LoDs of these areas can be useful for modelling practices in
terms of time and design simulations at lower or higher levels.
The modellative practices of Castel Masegra have been optimized trying to reduce
processing time, costs related to the generation of the model and increase its level of
interoperability for the next maturity level. Starting on the experience of the Basilica of
Collemaggio, it was useful to set a proper general LoDs. The large size of an HBIM
required ReverseLoDs practices to guide the extrados and intrados surfaces of the vaults.
The HBIM model of Visconti Bridge showed the morphological features of a medi‐
eval infrastructure bypassing the generative phase of 2D drawings, thanks to a LoD 500
supporting the entire decision-making process, facilitated the sharing of information in
different interdisciplinary groups for the protection and preservation of a historical
bridge. Shared knowledge phase provides a process of simplification of the model for
mobile devices and clouds in the web. The reversal definition ‘ReverseLoDs’ may
provide a selection of the information shared to improve the size of the model, the
navigation and its simulation.
9 Conclusion
Thanks to laser scanning technology for a modeler is easier to reconstruct the complexity
of the building. Simplifications of the model based on personal interpretive logic is still
mandatory of complex projects. This paper presents an advanced modeling technique
for historic buildings, which integrate the pure modelling and NURBS technology with
additional parametrization to orientate the models to the correct level of complexity.
The levels of development for the HBIM must ensure a bidirectionality in favor of
updates of the constructive elements of the building realized in digital environment for
different types of analyses. The ReversLoDs approach for historic buildings can support
the process of designing, building, restructuring, maintenance and analyses.
Finally, these techniques highlight the need to introduce a hierarchy of new levels
of accuracy of HBIM models giving a certified quality value of the generated model.
References
1. Azhar, S., Brown, J., Farooqui, R.: BIM-based sustainability analysis: an evaluation of
building performance analysis software. In: Proceedings of the 45th ASC Annual Conference,
Gainesville, Florida (2009)
2. Barazzetti, L.: Parametric as-built model generation of complex shapes from point clouds.
Adv. Eng. Inform. 30, 298–311 (2016)
3. Barazzetti, L., Banfi, F., Brumana, R., Gusmeroli, G., Oreni, D., Previtali, M., Roncoroni, F.,
Schiantarelli, G.: BIM from laser clouds and finite element analysis: combining structural
analysis and geometric complexity. In: International Archives of Photogrammetry, Remote
Sensing and Spatial Information Sciences, vol. 40(5/W4), pp. 345–350 (2015)
4. Barazzetti, L., Banfi, F., Brumana, R., Previtali, M.: Creation of parametric BIM objects from
point clouds using nurbs. Photogramm. Rec. 30(152), 339–362 (2015)
5. Brumana, R., Oreni, D., Cuca, B., Binda, L., Condoleo, P., Triggiani, M.: Strategy for
integrated surveying techniques finalized to interpretive models in a byzantine church,
Mesopotam, Albania. Int. J. Archit. Heritage 8, 886–924 (2014)
6. Eastman, C., Teicholz, P., Sacks, R., Liston, K.: BIM Handbook: A Guide to Building
Information Modeling for Owners, Managers, Designers, Engineers and Contractors, 2nd edn.
Wiley, Hoboken (2011). 626 pages
7. Fai, S., Graham, K., Duckworth, T., Wood, A.: Building information modelling and heritage
documentation. In: XXIII CIPA International Symposium, Prague, Czech Republic (2011)
8. Fai, S., Rafeiro, J.: Establishing an appropriate level of detail (LoD) for a building information
model (BIM) – West Block, Parliament Hill, Ottawa, Canada. ISPRS Ann. Photogramm.
Remote Sens. Spat. Inf. Sci. 2(5), 123–130 (2014)
9. Johnson, R.E., Clayton, M.J.: The impact of information technology in design and
construction: the owner’s perspective. Autom. Constr. 8(1), 3–14 (1998)
10. Oreni, D., Brumana, R., Banfi, F., Bertola, L., Barazzetti, L., Cuca, B., Previtali, M.,
Roncoroni, F.: Beyond crude 3D models: from point clouds to historical building information
modeling via NURBS. In: Ioannides, M., Magnenat-Thalmann, N., Fink, E., Žarnić, R., Yen,
A.-Y., Quak, E. (eds.) EuroMed 2014. LNCS, vol. 8740, pp. 166–175. Springer, Heidelberg
(2014). doi:10.1007/978-3-319-13695-0_16
11. Piegl, L.A., Tiller, W.: The NURBS Book. Monographs in Visual Communication. Springer,
Heidelberg (1997). 646 pages
12. Succar, B.: Building information modelling framework: a research and delivery foundation
for industry stakeholders. Autom. Costr. 3, 357–375 (2009)
13. Trani, M., Cassano, M., Todaro, B., Bossi, B.: BIM level of detail for construction site design.
In: 4th Creative Construction Conference, Krakow, Poland (2015)
14. Aravici, Y.: Towards building information modelling for existing structures. Struct. Surv.
26, 210–222 (2008)
15. Zhanga, J.P., Hub, Z.Z.: BIM- and 4D-based integrated solution of analysis and management
for conflicts and structural safety problems during construction: 1. Principles and
methodologies. Autom. Constr. 20(2), 155–166 (2011)
16. Zvietcovich, F., Castañeda, B., Perucchio, R.S., Morales, R., Uceda, S.: A methodology for
updating 3D solid models of complex monumental structures based on local point-based meshes.
In: 1st International Congress on Digital Heritage, DigitalHeritage, Marseille, France (2013)
Full Paper: Innovative Methods on Risk
Assesment, Monitoring and Protection of
Cultural Heritage
Applicability of 3DVE Indicator for Cultural
Heritage Landscape Management
Jaeyong Lee(&) and Youngmo Kim
Department of Heritage Conservation and Restoration,

Graduate School of Convergence Cultural Heritage,
Korea National University of Cultural Heritage,
Buyeo-gun, Chungcheongnam-do, Korea
{headoz,ymkim1683}@nuch.ac.kr
Abstract. Uniformly control of the height of buildings owned by individuals to

protect cultural heritage landscape causes social conflicts. Thus, it is necessary
to introduce an indicator that can simultaneously evaluate the criteria for control
of the height of buildings (CCBH) for urban development and cultural heritage
landscape management. We developed 3D Visual Exposure (3DVE) as a useful
indicator to review the validity of the CCBH around cultural heritage. By using
the 3DVE, it was possible to calculate visibility and we succeeded in mapping
the opportunity of view on 3D geospatial information and evaluating landscape
variation with statistics through changing building heights around cultural
heritage. We predict that the 3DVE presented in this study will have high
utilization as an indicator for the cultural heritage landscape management.
Keywords: Visible intensity evaluation Historic landscape conservation

Building height control Digital terrain model
1 Introduction
1.1 Background and Aims of the Study

Urban development and expansion form a contrasting relation against the protection of
cultural heritage. An urban management plan is established, using data on population,
atmosphere, transportation and facilities. Quantitative data are analyzed, and criteria are
prepared to control urban development capacities such as land use, building coverage
rate, building height, floor area ratio and green area ratio. In the meantime, a landscape
management plan for cultural heritage, too, begins with preparing criteria for control of
the height and usage of the surrounding buildings. In a city of which a cultural heritage
is its landmark, like St. Paul’s Cathedral, London, the CCBH are set up through view
corridor [4]. Most cities use a method of maintaining the line of sight to the cultural
heritage from a main view point.
On the other hand, in a radical or randomly developed city where cultural heritage
is not the landmark anymore, it is not easy to set up the CCBH for the cultural heritage
landscape management. Since 2007, in order to protect cultural heritage and harmonize
with the landscape of neighboring areas, the CCBH has been established by Cultural

DOI: 10.1007/978-3-319-48496-9_11
132 J. Lee and Y. Kim
Properties Protection Law in South Korea. The goals of the building height control are
to prevent cultural heritage from diminution, preserve skyline and background land-
scape and secure view (Fig. 1). However, no reasonable methods for the setting of the
CCBH have not been prepared, which may cause the excessive infringement of private
ownership of property.
Fig. 1. Purposes of the CCBH by cultural properties protection law in South Korea
In urban areas, social conflicts are caused by the overlapping regulations on

building height according to Urban Planning Law as well as Cultural Properties Pro-
tection Law. In order to solve this problem, it is necessary to measure changes in
landscape and prepare assessment methods. Especially, a quantitative landscape indi-
cator has high potential to be used as an important factor of urban planning along with
various development factors of the urban environment.
We attempted to develop a useful indicator in setting up the CCBH for urban
planning and cultural heritage landscape protection. For this purpose, we used the
3DVE to calculate visibility frequency quantitatively. We also evaluated changes in the
opportunity of view with the Pungnam Gate (Cultural Heritage) and its surrounding
areas in Jeonju, where the CCBH is applied redundantly. In addition, by mapping the
3DVE results of view point and cultural heritage including its surrounding buildings on
3D geospatial information, this study expressed changes in landscape and sought to
estimate statistics for comparison.
1.2 Related Works

From the urban planner’s point of view, the visibility issues become significant in
establishing policy [2, 3, 15, 19]. Many researchers have developed visual indicators
which human visual senses were taken into account focusing on city developments and
planning [16]. The process of the visibility analysis enables planners to identify
potential landscape character and its changes [12, 14]. Still, attempt to quantify the
amount of vision in architecture or urban space is actively in progress (Table 1).
Applicability of 3DVE Indicator for Cultural Heritage Landscape Management 133
Schroder et al. (2006) estimated the visible area two-dimensionally through an

analysis of the line of sight of the corners of the building. However, 3D topography has
not been considered. Yang et al. (2007) estimated viewed surface from a view point by
using buildings modeled with digital vector data. By measuring the visibility of the
building, the CCBH was proposed. Through simplified modeling of the building,
alternatives for control of building height were compared. However, like Schroder
et al., the topography of the actual study site was not reflected.
Bartie et al. (2010) analyzed the visibility of a wide range of areas by using data
surveyed with LiDAR. It was time-saving for a rough analysis of the visibility of a
wide range of areas. However, since the precision of the building data obtained from a
satellite was low, it did not materialize the shape of the building in detail, and there was
a limitation that alternatives according to changes in building height could not be
compared. It is noted that raster dataset based Visual Exposure is not suitable for urban
areas where built up structures are dominant [10–12].
Kim et al. (2015) developed a nurbs data-driven 3D visual analysis method and
proposed a method of overcoming the limitation of the visibility analysis by ArcView.
An et al. (2015) showed the usability of 3DVE indicator in quantifying the reduction of
the view chance in neighboring houses according to the construction of apartment
houses using Rhino 5.0 and Grasshopper.
Therefore, it is important to develop a vector data driven 3D visibility analysis
program in order to increase accuracy of the analysis [10, 11, 13] and it is need to
consider the possibility of the vector data based visibility analysis [5, 6].
Table 1. Comparison of related works

Previous Terrain Detailed Building Controlling Processing Work Cost
studies (elevation) building facade landscape speed efficiency
modelling modelling visibility variation
evaluation
Schroder X X X O Slow Low Low
et al.
(2006)
Yang X O O O Fast Low High
et al.
(2007)
Bartie O X X X Normal High High
et al.
(2010)
Kim et al. O O X O Fast Low Low
(2015)
An et al. O O O O Fast Low Low
(2015)
1.3 Fundamentals of 3DVE

Visual Access and Exposure (VAE) model is a method of measuring the amount of
vision, developed by Archea (1984) for the first time [1]. The VAE model is an indicator
that divides a space into the grid, adds up the frequency seen in the horizontal vision
from the level of human eyes or estimates the visibility from a certain view point.
We brought about this analysis and measurement method to a 3D urban space. The
3DVE is adding up the amounts of view point (road, square and park, etc.) from which
cultural heritage is visible within a certain range (Fig. 2(a)). Of course, a place with
high 3DVE can be one easily seen from the vicinities. By applying this estimation
method, we can find out the main parts of cultural heritage which can be seen best from
multi-view points (Fig. 2(a)) and the view point where cultural heritage can be seen
most (Fig. 2(b)). If you use this, you can find out the view point with excellent view of
the object.
Fig. 2. Calculation of 3DVE cultural heritage (a) and 3DVE view point (b)
Therefore, it is possible to review whether the view towards cultural heritage is

secured or not according to the changes of view point and visible areas of cultural
heritage. In addition, we can check skyline infringement and relative diminution of
cultural heritage against its surrounding buildings through a comparison of the visi-
bility intensity between cultural heritage and buildings in its background.
2 Methodology and Study Site

2.1 Process
The research was conducted in 4 steps.
In Step 1, the 3DVE analysis program was produced. It was developed with Visual
C++ so that it could operate in MS Windows. On the 3DVE, functions such as data
input, screen operation, analysis execution, analysis result checking and statistical data
storage were loaded.
In Step 2, a 3D DTM and building models were drawn up, using the data such as
topography, the height and the shape of buildings and cultural heritage in the study site.
For the drawing, a digital terrain map with the scale of 1/1,000 was used. As the
cultural heritage for which the visibility would be reviewed, the Pungnam Gate was
modeled. The Pungnam Gate (W45 × D32 × H16 m) was drawn up with mesh on
AutoCAD 2016. The maximum range to which the CCBH is applied by the Cultural
Properties Protection Law is within 500 m radius from the cultural heritage. Thus, the
area and buildings within 500 m radius from the Pungnam Gate were modeled. The
area is 1.4 km2, which includes a river and mountain. 4,479 buildings were modeled in
the form of a simple cube, ignoring the shape of the roof.
In Step 3, the CCBH were collected, which were applied to the study site according
to the Urban Planning Law and the Cultural Properties Protection Law. Referring to the
collected data, a present condition and two development scenarios (A and B) to
compare the landscape were prepared. Present Condition reproduces the current
building height and form (Fig. 3(a)). Scenario A is the landscape assuming the max-
imum development according to CCBH in the Urban Planning Law (Fig. 3(b)) while
Scenario B is the landscape assuming the maximum development according to CCBH
in the Cultural Properties Protection Law (Fig. 3(c)). At this time, the floor area of the
building was fixed, and the building height only was changed to adjust the develop-
ment capacity.
Fig. 3. Landscape modeling of present condition and two development scenarios
In Step 4, to use in a 3D analysis, a total of 1,769 target points1 were extracted from
the primary outline of the Pungnam Gate (Fig. 4(a)). 26,517 view points were selected
from road, square and park within 500 m radius from the cultural heritage (Fig. 4(b)).
From the surrounding buildings of the Pungnam Gate, 30,458, 89,538 and 61,520
target points were extracted from Present Condition, Scenarios A and B, respectively.
1
In order to calculate the visibility quickly, multiple ‘points’ should be selected from the important
parts of the object (We call this ‘target point’). Since no more than two sides of a building are seen
simultaneously, target points were calculated with 884, ½ of the 1,769 target points extracted.
Fig. 4. Examples of selected target points and view points
2.2 Value of Study Site and Landscape Control

The target of the research was the Pungnam Gate (Treasure No. 308) in Jeonju has a
relatively high importance among the cultural heritages designated by the state (Fig. 5
(a), (b)). The Pungnam Gate is the landmark of the city because it is a point-shaped
cultural heritage in which visibility as a target should be secured and also it takes
advantage of analyzing the view and its surrounding areas (Fig. 5(c)).
Fig. 5. Study site

The Pungnam Gate and its surrounding area have been controlled by the Urban
Planning Law. The area within 30 m radius around the Pungnam Gate was designated
as the Historic Cultural Aesthetic District, which is regulated at the maximum building
height less than 6 m. The main road in the east of the Pungnam Gate is the Center
Aesthetic District, which is regulated at the minimum building height of 9 m (3-story);
the 1st Class Residential District, at the building height of 15 m; and the 2nd Class
Residential District in the south, at the building height of 21 m. In addition, the other
areas are Commercial Districts, for which the building height is regulated to less than
21 m (Fig. 6(a)).
Fig. 6. Two CCBHs of the study site
On the other hand, the height of buildings in the area within 500 m radius around
the Pungnam Gate differ according to seven districts by the Cultural Properties Pro-
tection Law: District 1 is the whole area to the east within 30 m radius around the
Pungnam Gate, which is, in principle, for the preservation of the original topography,
and all actions of construction are subject to be reviewed by the Cultural Heritage
Committee. District 2 is the area within 50 m radius around the Pungnam Gate, in
which the maximum building height is regulated to less than 12 m (2-story) for the
buildings with pitched roof.
District 3 is the area within 200 m radius around the Pungnam Gate, in which the
maximum building height is regulated to less than 15 m (3-story). The regulations on
District 4 are the same as those on District 3. District 5 is the area extended to the north
more than 500 m radius around the Pungnam Gate, in which the maximum building
height is regulated to less than 18 m (4-story). District 6 is the area to the east with 300
to 500 m radius around the Pungnam Gate, to which the CCBH according to the nearby
cultural heritage is applied. District 7 is an area extended to the north-west south
towards more than 500 m radius around the Pungnam Gate, to which the level of
control of buildings is applied according to the Urban Planning Law (Fig. 6(b)).
3 3DVE Analysis Results

3.1 Changes of View Points
The 3DVE analysis result for view points is high around the square where the Pungnam
Gate is located and the whole area of the road traversing from east to west in three
landscape modellings. It is noted that the view opportunity of the Pungnam Gate is high
only on the axis of the road from east to west and from the square while there are
almost no points on the adjacent main roads from which it is seen (Fig. 7). As for the
3DVE of the road in Present Condition, 53.9 points (6.1 %) on average and up to 668
points (75.6 %) could be seen out of 884 target points of the Pungnam Gate (Table 2).
Fig. 7. 3DVE analysis results for view points
Table 2. Statistics of View Point 3DVE (left) and Result of Paired T-Test (right)
Present Scenario A Scenario B Assuming Equal Variances Scenario A Scenario B

std 141.7 139.7 140.0 Mean 46.865 47.064
avg 53.9(6.1%) 46.9(5.3%) 47.1(5.3%) Variance 19509.737 19595.583
max 668(75.6%) 668 668 Observations 26516 26516
Hypothesized Mean Diff. 19552.660
df 53030
t Stat -0.164
tCritcal two-tail 1.960
Significant 0.05
In the meantime, in the two scenarios, the 3DVE slightly decreases as compared to
Present Condition (Fig. 7(b), (c)). It is expected that similar landscape will be formed
in the two scenarios (As a result of a paired t-test, it was not possible to say that the two
were different at the level of significance of 95 %). It decreases by 46.9 and 47.1,
respectively, but there is no change in the maximum value. The 3DVE decreased
slightly, which means that the view from the road was not affected by the increase of
building height. In addition, the maximum value is maintained the same since there is
no change in the view from the vicinity of the square (Table 2).
Also, that there is no difference in the 3DVE in two scenarios means that the
CCBH, which was set up by the Cultural Properties Protection Law to protect cultural
heritage landscape, does not have an influence. In other words, even if the CCBH set up
by the Cultural Properties Protection Law are excluded and development by urban
planning is made, there is no impact on the view.
3.2 Visibility Intensity of Cultural Heritage (The Pungnam Gate)

As a result of the 3DVE analysis, in Present Condition, the Pungnam Gate was seen
from 341.6 view points on average (Table 3). The point best seen was seen from 2,749
view points. In the result of 3DVE mapping, the part best seen is the top of the roof to
the east of the Pungnam Gate (Left upper part in the Fig. 8).
Fig. 8. 3DVE analysis results for the north side of the Pungnam gate
Table 3. Statistics of the Pungnam Gate 3DVE (left) and Result of Paired T-Test (right)

std 502.6 456.7 457.0 Mean 292.698 293.8105
avg 341.6 292.5 293.6 Variance 208629.2 208917
max 2749 2276 2276 Observations 1768 1768
Hypothesized Mean Diff. 208773.1
df 3534
t Stat -0.0724
tCritcal two-tail 1.960 635
Significant 0.05
There is no meaningful difference in the 3DVE of the Pungnam Gate between the
two development scenarios (Fig. 8(b), (c)). The average decreases to 292.5 and 293.6,
respectively, and the maximum value decreases to 2,276, which is the same. There are
no differences in standard deviation (456.7 and 457) and mean (292.5 and 293.6) in
two scenarios. As a result of a paired t-test of the distribution value of the two sce-
narios, the similarity was very high (Table 3). This means that, even if the CCBH is
different, there is the same effect on preserving the view of the Pungnam Gate. In other
words, even if the heights of the surrounding buildings increase, the CCBH by the
Cultural Properties Protection Law does not affect practically. In addition, this suggests
that, even if the heights of buildings around the Pungnam Gate increase by the Urban
Planning Law, a situation which interfering the view of the cultural heritage does not
occur.
3.3 Visibility Intensity of Buildings Near Cultural Heritage

There was a tendency in the 3DVE of the buildings near the Pungnam Gate, very
different from that of cultural heritage or view points (Fig. 9). The 3DVE average of
the buildings around the Pungnam Gate was similar in Present Condition (7.6) and
Scenario B (8.4). The maximum value, too, shows similarity at 1,284 and 1,350,
respectively (Table 4).
Fig. 9. 3DVE analysis results for building in the area within 100 m radius around the Pungnam
gate
Table 4. Statistics of Buildings 3DVE (left) and Result of Paired T-Test (right)

std 53.3 80.1 61.6 Mean 7.54247 8.3555
avg 7.6 13.2 8.4 Variance 2792.715 3763.316
max 1284 1526 1350 Observations 30457 61519
df 69364
t Stat -2.079
tCritcal two-tail 1.960
Significant 0.05
As a result of a paired t-test assuming heterogeneous dispersion since the target

points differ as the building heights are different, it turned out that the 3DVE was not
the same in Present Condition and Scenario B at a significant of 95 %. In other words,
this means that, in spite of similarity, the two landscapes may have different views.
In the meantime, in Scenario A with large development capacity (Fig. 9(b)), the
3DVE of the surrounding buildings increases almost more than two times on average
from Present Condition (Table 4). The 3DVE result of the buildings near a cultural
heritage is high and this means that they may be seen better than the cultural heritage.
In other words, the possibility of weakening the identification of a cultural heritage can
be measured when the buildings form a landscape with it.
3.4 Skyline Changes of Cultural Heritage

The 3DVE result of Scenario B shows a noticeably lower distribution than that of
Scenario A and at the same time the distribution similar to the Present Condition, which
means that the CCBH by the Cultural Properties Protection Law is powerful. The
control of the surrounding building height affects the protection of the skyline with a
cultural heritage as background.
Fig. 10. Comparison of landscape simulations ((a)–(d): no. 1 north to south, (e)–(h): no. 2 south
to north)
In order to review the effect, landscape simulations were produced for comparison.
In the Fig. 10, buildings located at the back of the Pungnam Gate in Scenario A ((c)
and (g)) intrude the skyline while buildings located at the back of the Pungnam Gate in
Scenario B ((d) and (h)) does not intrude the skyline. Therefore, it is expected that the
3DVE will be able to evaluate the effect on the protection of the background and
skyline of cultural heritage.
4 Conclusions and Outlook
Controlling the capacity of urban planning properly is the main interest of urban
planning. In this situation, the landscape recognized as an important factor of the city
has not been treated carefully in the process of establishing a plan due to the difficulty
of measurement. With reaching the maximum capacity of urban planning, utilizing
simulation data that maintain the 3DVE to an appropriate level can contribute to
drawing effective measures for control of building height near cultural heritage.
Especially, cities holding cultural heritages have perceived the importance of the
landscape and set up a system to protect them; however, since no proper indicator for
measuring the landscape has been prepared, there was limitation in testing the relia-
bility and validity of their effects. The 3DVE proposed in this study showed a possi-
bility as an indicator that could assess the protection and infringement of the view of
cultural heritage and the intrusion of the skyline by the surrounding buildings.
A merit of the 3DVE is that it can intuitively judge the elements affecting landscape
by mapping the estimated results on the building elevation and the position of the view
point. In addition, it is expected that it would contribute to drawing the optimum plan,
comparing alternatives of the various CCBH by providing statistical analysis data.
References
1. Archea, J.C.: Visual access and exposure: an architectural basis for interpersonal behavior.
Ph.D. thesis, The Pennsylvania State University, Pennsylvania (1984)
2. Quintrand, P., Zoller, J., de Fillppo, R., Faure, S.: A model for the representation of urban
knowledge. Environ. Plan. 18(1), 71–83 (1991)
3. Rabie, J.: Towards the simulation of urban morphology. Environ. Plan. 18(1), 57–70 (1991)
4. Zacharias, J.: Preferences for view corridors through the urban environment. Landscape
Urban Plann. 43(4), 217–225 (1999)
5. O’Sulivan, D., Turner, A.: Visibility graphs and landscape visibility analysis. Int. J. Geogr.
Inf. Sci. 15(3), 221–237 (2001)
6. Turner, A., Doxa, M., O’Sullivan, D., Penn, A.: From isovists to visibility graphs: a
methodology for the analysis of architectural space. Environ. Plan. 28(1), 103–121 (2001)
7. Schroder, C. J.: Quantifying urban visibility using 3D space syntax: Technical report,
Unpublished MSc thesis, University of Edinburgh, Edinburgh (2006)
8. Yang, P.P., Putra, S.Y., Li, W.: Viewsphere-a GIS-based 3D visibility analysis for urban
design evaluation. Environ. Plan. B Plan. Des. 34(6), 971–992 (2007)
9. Bartie, P., Reitsma, F., Kingham, S., Mills, S.: Advancing visibility modelling algorithms for
urban environments. Environ. Urban Syst. 34(6), 518–531 (2010)
10. Domon, G.: Landscape as resource: consequences, challenges and opportunities for rural
development. Landscape Urban Plann. 100(1), 338–340 (2011)
11. Jorgensen, A.: Beyond the view: future directions in landscape aesthetics research.
Landscape Urban Plann. 100(1), 353–355 (2011)
12. Domingo-Santos, J., Villarán, R.F., Rapp-Arrarás, Í., Provens, E.C.: The visual exposure in
forest and rural landscapes: an algorithm and a GIS tool. Landscape Urban Plann. 101(1),
52–58 (2011)
13. Kirkpatrick, J.B., Daniels, G.D., Davison, A.: Temporal and spatial variation in garden and
street trees in six eastern australian cities. Landscape Urban Plann. 101(3), 244–252 (2011)
14. Park, B., Furuya, K., Kasetani, T., Takayama, N., Kagawa, T., Miyazaki, Y.: Relationship
between psychological responses and physical environments in forest settings. Landscape
Urban Plann. 102(1), 24–32 (2011)
15. Anastasopoulos, P., Islam, M., Perperidou, D., Karlaftis, M.: Hazard-based analysis of travel
distance in urban environments: longitudinal data approach. J. Urban Plann. Dev. 138(1),
53–61 (2012)
16. Zheng, Z., Bohong, Z.: Study on spatial structure of Yangtze river delta urban agglomeration
and its effects on urban and rural regions. J. Urban Plann. Dev. 138(1), 78–89 (2012)
17. An, K., Ko, H., Kim, C.: Development and implementation of visual exposure indicator for
residential development. Int. J. Smart Home 9(2), 1–10 (2015)
18. Kim, C., Ko, H., An, K.: Development and 3D visibility analysis models using NURBS in a
residential development. Int. J. Smart Home 9(1), 151–160 (2015)
19. Jie, H., Tsou, J.: GIS-based visual perception analysis of urban natural landscape for urban
planning supporting: a case study of Jinzishan Hill region. In: 19th eCAADe Conference
Proceedings, pp. 505–510. eCAADe Press, Helsinki (2001)
20. Schroder, C., Mackaness, W., Reitsma, F.: Quantifying urban visibility using 3D space
syntax. In: Geographical Information Science Research UK Conference, pp. 359–366.
GISRUK Press, Kildare (2007)
Taking the Next Step in Digital Documentation
of Historic Cities
How HERMeS Evolved in an Open Data Digital Library
of Historic Buildings
Pavlos Chatzigrigoriou(&)
Cyprus University of Technology, CY 3036 Limassol, Cyprus

[email protected]
Abstract. When a long-term research finishes, there is always a question about

implementation and further development. In the case of HER.M.e.S, the Digital
Heritage Management System of the Historic City of Hermoupolis, in a Greek
Aegean Island, Syros, it was also a question about raising awareness. The
research proved that 2.4 historic buildings collapse every year, as a result of
abandonment. This phenomenon was intensified by the severe economic crisis in
Greece. The research proposed an optimal conservation plan for the city, after
carefully evaluating variables through a multi-criteria model using GIS and an
innovating point system. But in order to apply this plan, we need to be able to
update the data, as buildings are constantly changing through time. Updating a
database with more than 1000 historic buildings, with no funds, is a huge
challenge. Soon, we decided that the only way to go is to use a crowdsourcing
method. Developing a digital heritage collection portal, using free open source
software and serving crucial data for every building, was the answer to our
problem. We asked citizens to check the data base, report mistakes, updates,
stories, photographs and use the portal to learn about their city. This effort led us
to a big digitization project, with up to 1290 historic buildings, 14.400
geo-tagged photos and more than 15.000 fields of information. The project
HERMeS, as a conservation plan and a heritage digitization project won the 2015
European Union Europa Nostra Award in Category Research & Digitalisation.
Keywords: Conservation Digital documentation GIS Heritage Buildings

pathology Historic city Open data
1 Introduction
Because of the severe economic crisis, Greek historic heritage is at risk. However,
historic cities like Hermoupolis, were dealing with this threat years before the crisis.
The current situation needs drastic action, with innovative low cost ideas. The main
objective of this research is to evaluate the vulnerability of historic buildings and to
propose a decision-making model that ranks the need of intervention in a historic
building stock. In order to achieve this, it is essential to design a digital database system
that enables the recording and indexing of the historic buildings pathology profile.

DOI: 10.1007/978-3-319-48496-9_12
Taking the Next Step in Digital Documentation of Historic Cities 145
Apart from pathology data, historic buildings are classified in terms of architecture
style and type, historic importance and spatial data (position in the city). Digital
monitoring building’s pathology profile has many positive impacts, such as knowing
the building’s conservation status, the need for intervention, and the risk of collapsing.
In a historic building stock, the safety assessment of buildings is a major research goal.
It needs a rational and quantitative evaluation, which usually involves six steps:
information acquisition, data processing, numerical modeling, evaluation of potential
scenarios, risk analysis, decision-making [1].
The city of Hermoupolis was chosen because of its uniqueness that can be traced on
its intact conservation almost as a hole until today. In Greece, there are only a few cities
that have managed to remain unharmed and even less are those that have endured the
unbearable pressures of “land granting”. In addition, the global interest for this city by
dozens of travellers for over 200 years, has maintained its physiognomy alive through a
number of literature works and manuscripts, which served as a protective shield for its
uniqueness. This protection endures until today in the consciousness of the inhabitants
as well as the visitors of Hermoupolis, and has produced many supporters of its
physiognomy ready to take action in case of its endangerment. Therefore, it should
come as no surprise that Hermoupolis presents a uniform image that cannot be found
anywhere else in Greece [2].
2 Theoretical Approach
Buildings express the society and people that constructed them. They are testimonies of
the cultural and social changes; the way and the time that buildings were designed,
constructed, used, abandoned and ultimately destroyed indicate the prime and decline
of a society. In that sense, the conservation of buildings as part of the cultural inher-
itance of a place is of crucial importance because it helps us arrive to many conclusions
about the development of our civilization. All cities may be characterized as large scale
“works of art”. However, this kind of art has a temporary nature, it cannot be controlled
and does not progress in the rhythm that other form of arts can, such as music for
example. In the course of time, various people may experience a city in totally different
ways [3]. Ingold, professor of Anthropology in the University of Aberdeen, mentions
in his book “The Perception of the Environment: Essays in livelihood, dwelling and
skill” that “building is a constantly continuing process, for as long as people inhabit a
certain space. It doesn’t have a clear starting point based on a design that we have
previously fashioned and it doesn’t conclude in a ready object. The “final form” is only
momentary….” [4]. Since human beings reside in a city only for a really brief period of
time compared to the lifespan of the city itself, they should not impose predetermined
schemes on it and its environment, rather, they should integrate to the city, develop it
according to the environment’s demands and inherit it to succeeding inhabitants.
Every facet of human activity has its Monuments charging them with important
historical memories. Therefore, Monuments may appear as formations in time
(buildings, cities, natural elements, or even music works), as points of reference (e.g.
March 25th, 1821), or even as prescribed behaviors (e.g. the New Year’s Gala in
Vienna). However, when it comes to structured environments, a monument has an even
146 P. Chatzigrigoriou
more essential role, which is to stimulate people’s memory and to function as a con-
necting link between the past and the future. Indeed, this role is dual, since the
monument provides answers to problems of the present, while retaining the memories
of the past. When the
monument is placed in
an urban area, it char-
ges memories, excites
imagination, and cre-
ates a certain mindset
to human beings [5, 6].
The monument as a
work of art is percep-
tible only in a specific
cultural milieu, not
everywhere, not
always, and certainly
not by everybody.
Heidegger uses the
word “Bewahrung”,
which means “conser-
vation”, but shares the
same root with the
word “Warheit”, which
means “truth”, denot-
ing in the term conser-
vation the conservation
of the “truth” of the
monument [7].
Buildings have
their own circle of life;
they get older, have
their own pathology,
are restored (some-
times poorly), aban-
doned and ultimately-
at the end of their life-
they collapse [8].
Buildings have three
different ages: The age
based on the date they
were built, the age
based on their materi-
als condition and the
age based on economic
and service factors [9].
The aging/deterioration Fig. 1. Pathology grades and building’s energy
process of buildings is a complicated procedure where a lot of research is taking place.

Samuel Harris in his book “Buildings Pathology” (pages 16–23) in order to describe
the deterioration mechanism, is viewing the building as an energy model and the he
uses the Second Law of Thermodynamics: Energy flows from higher (ordered) state to
a lower (random) state, increasing its “entropy”. Keeping in mind the First Law of
Thermodynamics (the building is a “closed” energy system), Harris is proposing that
the deterioration mechanism is inherently an exothermic process and the deterioration
is entropic (Fig. 1) [10].
To predict the life span of a building, one should be able to identify with certainty
all the variables involved in the deterioration model. Considering that it is simple to
identify the properties of a single material, one realizes the complexity of the issue
when it comes to buildings, where hundreds of different materials are involved. It is
also very likely in the same building to evolve two and more deterioration mechanisms
simultaneously, which further complicates the prediction process. Therefore, identifi-
cation of materials, their properties and damage-suffered buildings from the environ-
ment is the first step to approach a complex interdisciplinary problem as that of
determining the mechanism of deterioration and collapse of buildings. This matter
concerns the engineers of all disciplines.
With regard to the issue of rescuing interventions, building pathology as well as the
monitoring of building pathology, are of major importance. It’s a fact that when a
building is abandoned it runs into trouble. The first signs of humidity activate dozens of
wearing/erosion mechanisms on the building materials. Shortly after, they all start to rot
and collapse, one after the other. As soon as a building is categorized as “dangerous” in
the consciousness of the neighborhood tenants, it becomes a target for demolishment
[8]. This practice has been confirmed on many different occasions also in the present
study. Therefore, the objective is to monitor the pathology of buildings – monuments,
and to primarily help the ones that are in the process of being stigmatized as dangerous.
The use of the Information and Communication Technologies can be decisive
towards that direction. On a scientific level of the protection of architectural inheri-
tance, the contribution of information science was and continues to be of outmost
importance. It provides us with answers and solutions for the difficulties and problems
that arise during the evaluation process of the subordinate state of buildings, as well as
their monitoring and managing. From the relevant theoretical overview, it is concluded
that Planning Support Systems infiltrate remarkably slow mainly due to their limited
dissemination, their weakness in the choice of appropriate tools, their incomplete -for
their ability- update and the non-projection of good practices. This is a phenomenon
that can be traced in general in the distribution of new technologies [11, 12].
3 Methodology
First, a field research was conducted by a team of trained inspectors (civil engineers
and architects). Specifically, nine hundred and twenty-four (924) buildings were
recorded, creating inventories with one hundred and ninety-two (192) fields for each of
the examined building. The author trained the inspectors. The author answered any
questions or problems and the same guidelines were given to all inspectors, in order to
increase the credibility of the collected data. The field research took place in years 2004
and 2005. For this research, “recording” of a historic building is a holistic process in
order for the building to join the system HERMeS in a reliable and functional way.
This recording process was designed in a structured and simplified manner so as to
minimize the factors that may affect the reliability of the survey.
Before starting the recording process, the tool that was used for the required pur-
pose, has the form of a questionnaire. The design of the questionnaire is a very
important part of the development process of database systems HERMeS. This is
because the items included in the questionnaire are in fact the database fields and from
processing them one gets the research findings. Considering the above, a lot of thought
was spent for each section of the questionnaire in order to integrate important data for
the building, but with no unnecessary fields that could increase the recording time and
the size of the database. Special attention was given to pathology data recording, as this
is the primer research purpose. Every historic building was divided in sub-structural
units, considering what Greek and international bibliography propose [13–18]. In this
research a historic building is divided in ten (10) sub structural units (Fig. 2).
Recording pathology problems in a historic building is complicated, because it is
difficult to measure those problems. In order to simplify the process, a five (5) grade
scale was used in every pathology problem (Fig. 1): from grade “A” (it means that
there is no pathology problem) to grade “E” (crucial pathology problem). Pathology
data reliability is extremely important for developing a reliable conservation plan. This
is why there is a pathology profile picture for every single building (Fig. 2) and
HERMeS uses an algorithm to mass check all 924 buildings, for common errors during
pathology recording (Fig. 3). Regarding earthquake damages, the conservation state
documentation recorded all pathology data of the buildings, including -plausible-
earthquake cracks. Since HERMeS is applied in the same historic city, earthquake risk
is the same for all buildings. The variable of this risk is covered by the total pathology
grade of the building. Having said that, one can try multiple scenarios using only the
pathology grade of bearing walls, since they play a crucial role for complete failure
(collapse) of buildings during earthquakes.
The Database of HERMeS is originally designed according to user requirements
and needs of the present research. The main objective is to record all the pathology data
of “historic building”, architectural and historical data and any other information that
helps to protect the building in long-term. The data fed to HERMeS are spatial, which
vary both in space and time. Specifically, it is predicted in advance that the data
collected from evaluations of engineers in the field, will be associated with the Geo-
graphic Information System (GIS) HERMeS, and thus the database developed in to a
Geographic Database [19]. The spatial dimension serves many variables, especially
when assessing the risk of a building (e.g. location of the building in the city and
distance from a school).
The objective of the conservation of the physiognomy of Hermoupolis by pro-
tecting important building’s shells will not be achieved without ensuring mechanisms
slowing deterioration of buildings and making the necessary interventions to prevent
their collapse. In this respect, a key role has the GIS of HERMeS in combination with
the information provided by the database.
Fig. 2. Analyzing pathology profile in buildings, in order to increase the reliability of the data.
Fig. 3. An Algorithm checked all 924 buildings, for recording pathology errors.
By recording the pathology, HERMeS identifies the buildings at danger via an

appropriate algorithm. In the first application of the system those buildings were 160
from a total of 924 buildings. The prioritization of interventions in these buildings is

critical, as it is not possible to lower the collapsing risk simultaneously in 160 buildings
and the interventions cannot be judged solely by the reactions of local residents.
Bearing in mind the fact that one, given the current economic conditions, has to make
best use of the funds for this purpose, it is proved that the relevant decision requires
multi-criteria analysis method of prioritizing interventions. Specifically, the analysis
takes into account the risk of collapse for each building, but in connection with a series
of other variables, such as the role of building in Hermoupolis, the position in the city,
the influence in other areas of interest, the social impact etc. In the case study of
Hermoupolis the following variables were taken into consideration:
• Buildings’ proximity to schools
• Buildings’ proximity to touristic sites (harbor, museums, etc.)
• Buildings’ proximity to public buildings (municipalities, government offices etc.)
• Buildings’ proximity to city center
• The historic value of buildings (architectural importance)
• The lack of interest to restore buildings (by the owners)
• The pathology state of buildings
A team of experts was chosen to evaluate and classify the six (6) variables starting
with the most important variable and ending with the less important. The team con-
sisted of 30 scientists, architects and civil engineers, having a minimum of ten
(10) years of experience in the field of buildings’ conservation. Five (5) of the team
members were academics. The evaluation of the variables was conducted on-line via an
Internet survey process. The results were adjusted to percentages and then weighted on
a scale of 1 to 1000. The seventh variable (severity of building’s pathology) was
weighted equally with the average of the other six factors (143/1000) in order to avoid
chances of overruling.
4 Results
The multi-criteria model presented here is flexible. Buildings pathology grades are
changing constantly due to interventions and deterioration mechanisms. Schools,
touristic sites and other criteria are also subject to change. HERMeS is able to process
the updated data, in a fast and user-friendly way. Additionally, a tool in GIS was
developed to apply the multi-criteria model on the historic city building stock and
export an updated version of ranked risk assessment buildings.
Building’s pathology is a quality assessment factor, but it is not sufficient enough to
evaluate and rate building’s risk hazard. The above-proposed multi-criteria model
based on pathology indicators in combination with the architectural quality ones, as
well as with other social variables such as the area, the density, the usage of near
buildings, the tourist attraction, etc. can produce a hierarchy list of risk assessment
(Fig. 4). The point system leads to a Conservation Plan for the city of Hermoupolis,
giving the hierarchy of interventions that must be done in order to save the maximum
architecture heritage with the minimum funds, postponing the risk of collapsing. With
Fig. 4. Part of a GIS map; the threat of collapsing is measured with “risk assessment points”.
this in mind, the Municipality of Hermoupolis started applying this strategy as a

Conservation Plan.
The final resulted historic buildings’ ranking catalogue was proposed to the
municipality of Hermoupolis as the optimal strategy for protecting the physiognomy of
the city, with low cost and gentle maintenance techniques to be applied on the historic
buildings. It is noteworthy to indicate that the municipality has already adopted this
strategy and implemented on six (6) buildings. So far, the results have been encour-
aging since the deterioration process has been delayed and the risk hazard has been
reduced. These results validate the model developed in the present study.
Finally, the achievement of the documentation and management of important
building stock in Hermoupolis as implemented in this research, enrich the design of the
proposal for the inclusion of Hermoupolis in the list of monuments and sites of World
Heritage by UNESCO.
The research model has a huge social impact. For the first time in the history of
Hermoupolis, people are not afraid of their buildings. They know that there is a
monitoring system; they see that there is a conservation plan aiming to support specific
historic buildings. They also started to accept the historic value of buildings at risk,
giving information to the municipality in order to index them in HERMeS model. The
implementation of the research has an environmental value. Rescuing the Historic
Buildings at risk, we are reusing space and the city is not expanding. With HERMeS
model, we can evaluate the amount of space that is lost because of the lack of con-
servation and reuse of space (Fig. 5).
Fig. 5. The impact of research’s conservation plan on City expansion (environmental value).
The city in sixty (60) years, has expanded 110 % reducing the free area of the island. An island is
a “closed system”, surrounded by the sea. The free space is important as an environmental value.
Applying the Proposed Conservation Plan, we could reduce the need for expansion, because we
are saving historic building and we reuse the space. As shown at the graph, if the conservation
plan was used from 1945, Hermoupolis would be at today’s size, in 2026.
5 The Evolution
When this research ended, the author started thinking of how we will use all these
digital data in the future. It was a problem known in literature as “big data”, a term used
for collection of data sets so large and complex that is becomes difficult to process
using traditional data processing applications. The challenges include analysis, capture,
cu-ration, search, sharing, storage, transfer, visualization, and privacy violations. We
divided our data in smaller sets and we estimated what kind of information we can
release, who is going to benefit and what we expect from the data sharing. Then, we
researched the software more suitable for our project; a free, flexible and open source
web-publishing platform. We used OMEKA (www.omeka.org), an open-source soft-
ware platform developed by the Center for History and New Media at George Mason
University (Fig. 6). As a “web-based publishing platform for scholars, librarians,
archivists, museum professionals, educators, and cultural enthusiasts” (Omeka About
Page, 2009), Omeka resembles WordPress, with its plugins and themes, but is a very
different platform.
Fig. 6. Omeka is a free open source software for digital collections. Thinking “out of the box”,
we verged HERMeS not as a Conservation Plan any more, but as a big data digital collection. It
was the only way to ensure awareness, updating and enrichment for our Database. The Omeka
free software provides also the option of “Omeka.net” a low budget hosting. We choose this plan
in order to minimize the demands for support and web hosting. HERMeS is currently the biggest
online historic buildings digital collection. It has 5.550 unique visitors and 5–20 visits per day.
Architects, engineers, citizens of Hermoupolis, tourist offices and tourists use it every day for
information. The contributions are low, but after the Europa Nostra Award we predict a deeper
evolvement of local habitants.
Omeka is OAI-PMH compliant, and is based on Dublin Core. It can ingest

metadata in both CSV format and in XML. Point it at open access repository and it can
import collections whose XML is exposed for harvesting. The author developed a web
site hermoupolis.omeka.net using omeka and omeka.net hosting. More than 14.000
High Definition geotagged photos were uploaded, 1.290 Historic Buildings (and
adding more), more than 20 core data fields per building, summing up to 36.000
information objects. The buildings were tagged with info such as pathology grades,
architecture style, usage, size, historic value, protection laws etc.
Adding old photos (digitized) of these Historic Buildings was crucial for the
Historic Value of this project. We used photos from 1995, 2005 and 2014 (and adding
more). Researchers, Architects and Students can search, discover and learn about
Neoclassicism, an important Architectural Period in Europe, from this Digitization
Project.
The metadata are following International Standards and most of the information is
in English, in order to accommodate foreign users and Europeans.
Now, HERMeS is a complicated scientific system with two unique and prototype
pillars:
• It measures the pathology and estimates social factors in order to provide an
innovate point system to all historic buildings at risk. This point system is the base
of the conservation plan
• It deliberately shares the most important fields (data) of 1.290 historic buildings, in
a free open source web platform, providing useful information and asking peoples
feedback, in order to update its data.
Because of the above approach to the complicated problem of Historic City Con-
servation, in 2015 HERMeS is honored with the European Union Europa Nostra
Award, in the category of “research and digitization”.
6 Conclusions – Future Work
This Digitization Project is based on a research for a solid conservation plan for the
Historic City of Hermoupolis. Serving historic “Big Data” online makes this project
special. It adds to the Historic Value of Europe, because it presents live (online) and
organized an important part of European History; the Neoclassicism, as a movement
with all its aspects, in one city. The goal of this digital preservation is the accurate
rendering of authenticated content over time. At the same time, we still update our
HERMeS data, preventing the ageing of the Historic Buildings, following our bigger
Conservation Plan.
People are using HERMeS for different purposes:
(1) Scholars, use HERMeS to learn, for an essay or for a dissertation, to share
collections and collaborate with others in the creation of a project for scholarships.
(2) Educators use HERMeS to build inquiry-based tasks for students, create lesson
plans or build learning modules with their team.
(3) Architects use HERMeS to study Neoclassicism, to understand the different

styles, to find out the history of specific buildings and discover old photos.
(4) Engineers use HERMeS to understand Pathology, to find out the Pathology Grade
of their project and to learn about the past of Historic Buildings.
(5) Citizens use HERMeS to find out useful information about their buildings, to
learn about the pathology grades and of course to contribute stories for Historic
Buildings.
At the end, HERMeS is a Digitization Project of Cultural Heritage with a double
conservation aim: (i) Digital Preservation of the History (tangible and intangible) and
(ii) Conservation of Historic Buildings. This double aiming (a “Lighter” Portal and a
“Heavy” Scientific System) makes this project unique and challenging.
In the future HERMeS can adopt the know-how from 4D-CH world project and
develop more in the “timeline” of historic buildings [20, 21]. Using technology for
creating 3D reconstructed models and applying the digital data in an engaging way is
one of the goals. Another goal is to connect tangible and intangible heritage [22], a
challenge that HERMeS is already answering through a postdoc research.
Acknowledgments. 4DCH (http://www.4d-ch-world.eu/) project has received funding from the

European Union’s Seventh Framework program for research, technological development and
demonstration under GA no 324523.
EU Prize for Cultural Heritage/Europa Nostra Awards (http://www.europanostra.org) was
launched in 2002 by the European Commission and has been organised by Europa Nostra since
then. The Prize celebrates and promotes best practices related to heritage conservation, man-
agement, research, education and communication. In this way, it contributes to a stronger public
recognition of cultural heritage as a strategic resource for Europe’s society and economy.
References
1. Zonta, D., Pozzi, M., Zannon, P.: Managing the historical heritage using distributed
technologies. Int. J. Archit. Heritage 3, 200–225 (2008)
2. Travlos, I., Kokkou, A.: Hermoupolis. Commercial Bank of Greece, Athens (1980)
3. Lynch, K.: The Image of the City. MIT Press, Cambridge (1960)
4. Ingold, T.: The perception of the environment: essays in livelihood, dwelling and skill.
Routledge, London and New York (2000)
5. Stefanou, J.: The concept of the monument and the ideology of protection, Athens.
University publishing, N.T.U.A. (2004)
6. Papalexopoulos, D., Kalafati, E., Papadopoulos, S.: Building memory. In: Proceedings of the
2001 Conference on Virtual Reality, Archaeology, and Cultural Heritage, VAST 2001,
pp. 27–32. ACM, New York (2001)
7. Lefas, P.: Architecture and Inhabitation: From Heidegger to Koolhaas. Plethron, Athens
(2008)
8. Brand, S.: How Buildings Learn: What Happens After They Are Built. Penguin Books, USA
(1995)
9. Levy, M., Salvadori, M.: Why Buildings Fall Down. WW Norton & Company, New York
(2002)
10. Harris, S.: Building Pathology, Detioration, Diagnostics and Intervantion. Wiley, New York
(2001)
11. Darzentas, D., Darzenta, J., Spyroy, T.: Choosing appropriate tools by means of intelligent
decision support. In: JDS, pp. 1–17 (1997)
12. Vonk, G., Geertman, S., Schot, P.: Usage of planning support systems. In: Leeuwen, J.P.V.,
Timmermans, H.J.P. (eds.) Design & Decision Support Systmes in Architecture and Urban
Planning. Springer, The Netherlands (2006)
13. Antonakakis, D., Tsiomis, I.: Conservation study for restoration of the historic center of
Hermoupolis, Athens (1996)
14. Stefanou, J.: The Urban Planning of Hermoupolis: A Pilot Urban Study of a Historic City.
National Technical University, Athens (2003)
15. Organization for Anti Seismic Planning and Protection. Checking Structural Venerability,
Ministry of Environment. http://www.oasp.gr/node/76
16. Spanos, Ch., Spithakis, M., Trezos, K.: Methods for On-Site Valuation of Material’s
Characteristics. TEE, Athens (2006)
17. Watt, D.: Building Pathology. Blackwell, Oxford (2007)
18. Thornes, R., Bold, J.: Documenting the Cultural Heritage. Getty Information Institute,
Los Angeles (1998). European Foundation for Heritage Skills
19. Koutsopoulos, K.: Geographic Information Systems. Papasotiriou Press, Athens (2005)
of tangible cultural heritage objects from web-retrieved images. Int. J. Herit. Digit. Era. 3,
431–451 (2014)
21. Ioannides, M., Hadjiprocopi, A., Doulamis, N., Doulamis, A., Protopapadakis, E.,
Makantasis, K., Santos, P., Fellner, D., Stork, A., Balet, O., Julien, M., Weinlinger, G.,
Johnson, P. S., Klein, M., Fritsch, D.: Online 4D reconstruction using multi-images available
under open access. ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci. II-5/W1, 169–
174 (2013). doi:10.5194/isprsannals-II-5-W1-169-2013
22. Cacciotti, R., Valach, J., Kuneš, P., Čerňanský, M., Blaško, M., Křemen, P.: Monument
damage information system (Mondis): an ontological approach to cultural heritage
documentation. In: ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial
Information Sciences, vol. II-5/W1, 2013 XXIV International CIPA Symposium, 2–6
September 2013, Strasbourg, France
Risk Analysis and Vulnerability
Assessment of Archeological
Areas for the Preventive Conservation
Giuliana Quattrone(&)
National Council of Research CNR, Institute of Atmospheric Pollution Research,

UNICAL-Polifunzionale, 87036 Rende, Italy
[email protected]
Abstract. The territorial structure of archaeological sites is often compromised

by a variety of factors which, over time, can contribute to aggravate the dete-
riorating conditions of the archaeological areas. Natural disasters, often, have
caused irreversible damage (crashes, loss of finds, etc.) at the archeological sites.
The paper shows the results of research aimed at developing an innovative risk
assessment model, oriented to sustainability criteria, that can provide help in
making decisions process about the protection, conservation and valorisation of
archaeological areas. The proposed work presents a logical and operative model
suitable to estimate the actual risk condition for archeological sites developed by
integrating the information concerning vulnerability of archeological areas and
the danger condition of the sites upon which they are located. The determination
of the risk, derived from the evaluation of potential co-presence of independent
sources of danger within a given geographical area, has been modeled by a type
of approach multi hazard.
The methodology is applied at the case study of archaeological area of Sybari
in Calabria Region, in south of Italy, where tried to develop an integrated and
coordinated cognitive methodology - evaluation within an overall information
system for assessing the vulnerability of archaeological heritage and environ-
mental dangerousness in order to identify the level of risk which is under the
archaeological area in relation to the territory of reference.
The results obtained allow to relate the risk with the risk factors, the detection
of damage on archaeological heritage in question, determine the tolerance
threshold within which it is possible the preservation of goods and to propose a
synthesis strategy between environmental protection and sustainable
exploitation.
Keywords: Archaeological areas Risk management Vulnerability

Territorial planning Resilience Environmental sustainability
1 Introduction
The paper describe the results of the research aimed to establish the level of risk of an
archaeological resource in a particular area (geo-referenced), applying a comparative
evaluation based on the characteristics of the goods valued in relation to the reference

DOI: 10.1007/978-3-319-48496-9_13
158 G. Quattrone
context, compared with the degree of dangerousness of the territory, on the basis of a
shared taxonomy territorial and a preventive spatial assessment designed to measure the
definition of the threshold of tolerance within which it is possible the recovery and
preservation of goods and cultural areas also with respect to their environmental
sustainability.
The analysis of the degree of conservation/degradation of the natural heritage and
environmental components, is conducted considering interactions respect to a quali-
tative measurement scale and depending on some variables: the probability, the
duration, the frequency and the reversibility of impacts.
The research work is based on methodological approaches that seek to develop
innovative methods of analysis and risk assessment to improve the effectiveness of
mitigation actions and management in the territory and to improve vulnerability
assessment and risk level to which the cultural heritage is subjected in relation to the
territorial context y of reference. In this context, the archaeological heritage of the area
of Sybaris, in the first instance, was known and subsequently assessed, using several
criteria to recognize the cultural value and to assess the risk in an ordinal and com-
parative ranking (rank order). This rank was achieved through multi-dimensional
evaluations [8].
The demonstration of the level of risk is related to the phases of the “life cycle” of
the urban area and the territory in general, and also is related to the type and quality of
settlement of historic centers, as well as the recognition of the specific categories of
obsolescence that determine the asset depreciation (the loss of intrinsic economic value
and value in use). The risk is analyzed according to three profiles: the probability that a
risk event to occur; the consequences if the risk event occurs; the necessary measures to
prevent or minimize the expected consequences [8].
The research work was carried out within the SI.MO.NA Project, “Systems and
technologies for the Monitoring of cultural areas in the marine and terrestrial envi-
ronment” with financial supporter of the Operational Regional Project of Calabria
Region, FERS 2007/2013 and with technical supporter of Institute of Atmospheric
Pollution Research and Culture and innovation consortium of Calabria Region.
2 Vulnerability and Danger in the Archaeological Areas
The local risk is measured using a model that takes into account the local variation
range of dangerousness present in a given territory and the state of conservation of the
properties examined. The overall risk derives from valuation of the potential
co-presence, within a given geographical area, of independent sources of dangerous-
ness. In order to identify which parts of the territory to secure for to protect the integrity
of the existing cultural heritage it is necessary to determine the local risk associated
with each asset and also the global risk [4].
The preventive estimate of the risk, the prevention of damage and conse-lowing
reduction of possible degradation factors on capital, are based on a knowledge of the
environmental and territorial phenomena and assets to be protected.
As an alternative to the restoration which occurs at the expense occurred, this
address proposes to develop, through the intervention of conservation and maintenance
Risk Analysis and Vulnerability Assessment of Archeological Areas 159
of the goods, a strategy based on prevention of the damage from the prevention of the
degradation process through the environmental control and land control [5].
The model developed in the research and presented in this paper is a logical model
which takes account the local variation of the dangers present in a given territory and
the state of conservation of the properties examined. Appropriately crossing the hazard
data with vulnerability data of cultural heritage, it is possible to calculate, for each asset
located in a specific area of the territory, the value of the Global Risk associated with
the good as a whole. This risk dependes from the assessment of the potential
co-presence, within a given geographical area, of independent sources of danger.
The determination of the model has been shaped by a multi hazard approach or
“cross-cutting risks”. the risk is the product of two terms: hazard and vulnerability
R = VP ð1Þ
The vulnerability (V) of the cultural property, is a function that indicates the level of
exposure of a given good to aggression of territorial environmental factors, the terri-
torial hazard (P), is a function that indicates the level of potential pressure characteristic
of a given territorial area, regardless of the presence or not of the goods [4].
Two other variables are also considered:
“Resilience”, defined as the ability, refers to any body, individual or organization,
to cope and to recover from the effect of perturbing action produced by a negative
event, which depends on the capacity of adaptation and intrinsic regeneration to the
system;
The “Resistance”, is the attitude of imperturbability of system [7].
Vulnerability and resilience depend on the interaction of socio-economic elements
and fragility and ecosystem elements and on political and institutional determinants of
land (aspects not yet systematically covered in the construction of synthetic indexes).
Dimensioning the model, territorial unit corresponds to the wide area conterminous to
the archaeological area, or a good part of the municipal area.
Through an Overlay Mapping method, the model seeks to relate the hazards factors
existing on the territory and the conditions of the archaeological area vulnerability,
trying to arrive at a definition and quantification of the objective factors of hazardous
agents on individual assets. Obviously assuming the municipal area as the territorial
reference unit will have an effect of generalization of the “dangerousness values”
modeled due to the fact of heterogeneity of sources and to the different factors hazard
detection systems [12]. The risk is determined by this model using a multi hazard
approach (or “cross-cutting risks” or “interdisciplinary risks”) taking into account a
large number of variables and interaction of potentials independent events. In this
sense, the developed model allows to account of how effectively the different localized
contexts affect the actual state of risk to which an asset is subjected, from its current
state of conservation.
The risk is considered to be a function of two different sizes, the danger that is the
presence or likelihood of damaging events, and the vulnerability, that is the attitude of
the asset to be damaged [1].
160 G. Quattrone
The vulnerability of an asset is concretely manifested in his state of conservation: it

is therefore a measurable dimension through the various aspects of degradation. More a
good is wasted, more it is vulnerable to the aggression of the environment [7].
3 The Study Area
The archaeological site of Sybaris is located on the Ionian coast of Calabria, at a short
distance from the mouth of the Crati river. This part of the Calabrian territory, topo-
graphically known as Sybaritid, saw the birth, expansion an decline of the great poleis
of Sybaris. After the destruction of the Greek town, the Hellenist centre of Thurii first
and then the Roman town called Copia were built on its ruins. Such exceptional
stratification made Sybaris one of the widest and most important Archaic and Classical
Age archaeological sites of the Mediterranean [6].
Fig. 1. Locational scheme of archeological area of Sibari
The history of the site starts in 720 B.C. with the foundation of the Achaean colony
of Sybaris. In 510 B.C. the town was destroyed by Croton but was rebuilt in 444–443
B.C. as the Panhellenic city of Thurii. In 193 B.C. the Roman colony of Copia was
founded on that same site; it was turned into a Roman “municipium” in 84 B.C.
During the Imperial age, from the 1st to the 3rd century A.D., the town knew a
period of prosperity, but the marshy nature of the land led to a slow decay and gradual
abandon in the 5th and 6th century A.D. The site was abandoned definitively in the 7th
century [6].
For the taxonomy of resources were used the data of filing of the assets of the
Archaeological Superintendence of Calabria and the results of excavation carried out
by the Italian Archaeological School of Athens in 2013.
Starting from the 19th century, the growing interest for the still unknown and
mysterious history of Sybaris allowed to locate the site of the ancient settlement and
finally start the excavations in the ′30s, after several decades of investigations based
solely on the topographical directions provided by historical sources. The great
archaeological campaigns of the ′60s and those of the following decades, however,
only allowed to explore a small part of the large area of Sybaris, mostly unearthing
ruins connected to Copia and the Roman Age [6].
The area of the archaeological park covers 168 hectares and is divided into sectors,
each of which is identified with the name of the excavations: “Parco del Cavallo”,
“Prolungamento strada”, “Casabianca” and “Stombi”.
With specific reference to the Archaeological Park area, research has highlighted a
subsidence phenomenon evidenced by the presence of three overlaid cities: Sybaris
(720–510 BC), Thurii (444–285 BC) and Copiae (193 BC). The rate of subsidence has
gone reducing, but in recent decades the rate of subsidence has also become important
for anthropogenic causes, due to the archaeological excavations and the deeper aquifer
withdrawals, particularly in the levels high permeability below. The several problems
concerning the excavations in the area the alluvial layers covering the ruins, the
overlapping foundations of the three towns, the water table hindering the archaeo-
logical operations required an accurate planning of the archaeological interventions.
4 Problems Related to the Degradation and Management

of the Archaeological Areas
The flooding of the Crati, due to the breaking of the banks, a few years ago, did arrive
in the archaeological area about 200 thousand cubic meters of water and debris that
have flooded completely five hectares provoking several damage. For the mud we have
Fig. 2. Plan of the archeological excavations

162 G. Quattrone
had to clean up the street and, above all, to clean and restore the area of the theater, of
the Thermae and the Roman houses.
The structures of the various areas of Sybaris Archaeological Park, are located at a
lower level to the level of the aquifer water and therefore it is need a pump system that
provides the lowering of the aquifer water under the floor of the structures [10]. The
well-points system, active 24/24 h and whose management is contracted out to external
companies, is subject to occasional malfunctions, influenced by weather conditions,
which cause periodic rises of the water level inside the archeological area.
Fig. 3. The archeological area of Sibary after the flooding of the Crati river
The structures are therefore subjected to a stress due to the presence of water. Two
are the most immediate consequences caused by humidity the growth of vegetation
which with its own root system damages the structures; the erosion of incoherent soil,
on which rest the old structures [9].
Some interventions, completed in 2012, have made a significant contribution to
minimize humidity damage:
– Were made same metal fences containing soil subject to erosion underliving to the
mausoleums of the Roman period, fate, in fact, on top of a layer of soil accumulated
over the Hellenistic structures;
– Are punctual works were started filling in depressed areas without archaeological
structures, reducing the areas where water stagnates so as to inhibit the growth of
wetland vegetation.
The amount of filling and leveling works with soil are not yet concluded and it is
necessary to provide for additional works of fill with soil to eliminate the remaining
areas subject to the waterlogging.
A separate problem concerns the degradation of limestone blocks of Hellenistic

structures, mainly due to phenomena of degradation of the stone disintegration,
accentuated by repeated cycles of wet and dry, which inevitably affects the structures.
In addition, the pre-Hellenistic walls feel considerable damage due to the disintegration
caused by the action of atmospheric agents; the eastern sector of the area, is often
occupied by water due to the inability of the current system to ensure a continuous and
effective drainage [11].
Many of the ancient walls, finally, require integrations and completions, with
adequate restorations, to facilitate the readability of the architectural complex.
5 The Risks Identification
The risk factors considered in the model are several and cover the aspects more
specifically related to environmental risk, the aspects related to the structural risk, those
related to the risk due to the use, the aspects related to the risk due to malicious acts, or,
more generally, to anthropic risk or economic risk.
The assessment was developed through the following phases:
– Analysis: Identification of important periods from an archaeological and historical
point of view, regarding the geographical area considered.
– Sensitivity: Definition qualitative/quantitative sensitivity of the historic period.
– Risk Assessment: definition qualitative/quantitative of the risk level.
This is a type of probabilistic and presumptive evaluation having to define ex ante
the territorial mutations, over time.
From a theoretical and methodological point of view, the “value”, which can be
identified by term “sensitivity” (S), of a component, is the result of the product of the
fragility (f) inherent in the component, and its vulnerability (v).
S = f:v ð2Þ
Where the fragility is the characteristic of a component and the vulnerability is the
probability that this component can be “assalted” by an external pressure source [1].
The indicators used to define the archaeological sensitivity are: Uniqueness, Rarity,
Antiquity, Conservation Status, Artistic Merit.
The characteristics of each indicator were represented through four hierarchical
levels of quality, which is assigned a numeric value from 1 to 4, except for the
Uniqueness indicator, which may be present or absent.
The vulnerability to the archaeological areas is represented by a global index
expressed by a growing number from 0 to 10, where increasing of the index corre-
sponds the higher vulnerability. The vulnerability was calculated, statistically, on a
significant number of variables that precisely describe the state of preservation of the
archaeological area.
The acquisition of information about the vulnerability of archaeological heritage
took place through a phase of cataloging that has provided a metric assessment of
individual elements that can be each real estate (foundations, high, slab structures,
164 G. Quattrone
Fig. 4. Sensitivity matrix
stairs and vertical connections, covers, floors, walls) present in the archaeological area,
the forms of alteration observed, and an estimated of the extension and severity of the
degradation [3].
In the study of Sybaris area we tried to analyze every constructive and decorative
element, of the area of domus, of the are of excavation of the Cavallo Park by relating
the damage found to the techniques and materials that constitute them. A same form of
alteration, in fact, can take a meaning and a different valence depending on the material
used and the manner in which this is put in place in the structures. The mechanical
characteristics, the shape, the size and the material constitute further elements of
assessment in the analysis of archaeological vulnerability.
For to calculate the vulnerability also we have tried to collect the information on
other specific aspects. In many cases, the possibility of conservation of the ruins is not
only a function of the conditions of the physical characteristics of the materials, but is
also determined by a number of factors, extrinsic to its physicality, but who have a
direct feedback on its possible decline. In other words the judgment on the status of an
archaeological monument conservation, not only based on the semiotics of traditional
damage, must also occur to verify:
– the presence, the effectiveness, the adequacy and the maintenance of the architec-
tural protection artifacts;
– the existence and the effectiveness of out flow systems and drainage of rain waters
devices;
– the presence, the effectiveness and the adequacy of the partial or seasonal protection
measures (e.g. the waterproofing of walls filled or the floor covers during the fall
and winter);
– the existence of a routine maintenance program and weeding;
– the presence and the efficiency of security and surveillance devices (such as the
guard, the fences, etc.),
– the compatibility of the possible uses of the ancient structures.
To calculate the vulnerability indicator, a statistical - descriptive approach was
used, in which each considered information is fed back to a limited number of sortable
categories according to a scale of increasing value. The model for the calculation has
been set considering each archaeological resource analyzed as a statistical unit
“complete” regardless of the number and the type of the elements that constitute it.
Information processed are those arising from the analysis of the seven constitutive
and decorative elements, implemented, taking as additional parameters, the following
categories:
– Use,
– maintenance,
– security systems,
– Installations water drainage system
– Disposal system water/water-scooping,
– Protective covers
The set of the parameters described contributes to the calculation of the vulnera-
bility index and hazard, but each information intervenes in different measure, on the
basis of an appropriate “weights” system. The state of conservation of the building
blocks is given more weight than the other considered variables (use, maintenance,
protection systems, installations). But even among these there are differences that have
led to a hierarchy of weights. The roof structures are analyzed, also the architectural
structure in the individual components of the artifact are analized by evaluating for
each the adequacy and efficiency in relation also to the type of the various structural
parts and materials used.
The information so collected are properly reprocessed and evaluated according to
specific qualitative relations with the aid of statistical methods.
The sensitivity level for the archaeological site of Sybaris is placed in a
medium-high range.
From the reading of the environmental analysis we can draw the hazard matrix.
With the level of risk means the likelihood that the risks could interfere, generating a
negative impact on the presence of objects and artifacts airborn present in the
archaeological area. It can define the level of risk within a range from 0 to 3, or:
– Risk Null (numeric value 0).
– Risk Low (numerical value 1).
166 G. Quattrone
High Medium Low Null

Presence of the risk
X
Hydrogeological Risk
X
Erosion Risk
Pollution Risk X
Earthquake Risk X
Desertification Risk X
Fire Risk X
Subsidence Risk X
Sinkholes Risk X
Environmental Risk X
Human Pressure Risk X
Total Risk 15 6 1 0
Fig. 5. Hazard matrix
– Medium Risk (numerical value 2).

– High Risk (numeric value 3).
The level of risk to the archaeological site of Sybaris is 22 that is placed in the high
range.
At this point, pondering of the archaeological component, performed through the
definition of the environmental sensitivity, on the basis of the discoveries and infor-
mation in the literature, assessing the value of different historical periods, in compar-
ative way, and risk identification, such as probabilistic factor, capable of generating a
negative impact on the presence of objects and artefacts of archaeological interest allow
us to proceed to the final evaluation.
Assessment is carried out through the use of a matrix that calculate the total
cumulative risk that you might have on the archaeological site and define preventive
actions to be activated in the archaeological field.
6 Conclusions
The inferential evaluation systems for multi-dimensional estimation (Inferrential Evalua-

tion Engine) of cultural heritage (archaeological areas) are, in many respects, very
important for the current and future compatible use of cultural heritage, especially if
managed with geographic software inferential, as well as methods of support to the
decisions, because they allow to switch from a simultaneous vision to a synthetic vision [2].
However the intervention and enhancement choices they must carried out starting
from the type of quantitative/qualitative considerations mainly related to the level of
deterioration of the asset beyond the reference context. The demonstration of the level
of risk is related to the phases of the “life cycle” of the urban area and the territory in
general, to the type and quality of settlement of historic centers, and to the recognition
of the specific categories of obsolescence that determine the asset depreciation. Starting
from the case study of the archeological area of Sibari has been developed an inno-
vative risk assessment model, by a type of approach multi hazard, oriented to sus-
tainability criteria, that can provide help in making decisions process about the
protection, conservation and valorisation of archaeological areas. The model allows to
estimate the actual risk condition for archeological sites through integrating informa-
tions concerning vulnerability of archeological areas and the danger condition of the
sites upon which they are located. The determination of the risk, derived from the
evaluation of potential co-presence of independent sources of danger within a given
geographical area within an overall information system for assessing the vulnerability
of archaeological heritage and environmental dangerousness in order to identify the
level of risk which is under the archaeological area in relation to the territory of
reference. The results obtained allow to relate the risk with the risk factors, the
detection of damage on archaeological heritage in question, determine the tolerance
threshold within which it is possible the preservation of goods and to propose a
synthesis strategy between environmental protection and sustainable exploitation.
References
1. Abhas, K.J.: Cultural heritage conservation. In: World Bank, Safer Homes, Stronger
Communities: A Handbook for Reconstructing after Natural Disasters, Washington,
pp. 173–79 (2010)
2. Canuti, P., Margottini, C., Fanti, R., Bromhead, E.N.: Cultural heritage and landslides:
research for risk prevention and conservation. In: Sassa, K., Canuti, P. (eds.)
Landslides-Disaster Risk Reduction. Springer, Berlin (2009)
3. Campeol, G., Pizzinato, C.: Metodologia per la valutazione dell’impatto archeologico.
Archeologia e Calcolatori 18, 273–292 (2007)
4. Corcelli, A., Ioannilli, M., Cacace, C.: Progettazione e sperimentazione di un modello di
analisi Multi–hazard per la valutazione del rischio locale dei beni culturali. In: Atti della 12°
Conferenza Nazionale ASITA, pp. 839–846 (2008)
5. Ferroni, A., Cacace, C.: Carta del rischio: la vulnerabilità archeologica. http://www.icr.
beniculturali.it/documenti/allegati/Vulnerabilità_archeologica
168 G. Quattrone
6. Ministero dei beni e attività culturali - Sopraintendenza Archeologica della Calabria. http://
www.archeocalabria.beniculturali.it/archeovirtualtour/calabriaweb/sitoparco.htm
7. Graziano, P.: Rischio, vulnerabilità e resilienza territoriale. In: Quaderni del Dipartimento di
scienze economiche e sociali, dell’Università Cattolica del Sacro Cuore di Piacenza, serie
rossa economia, quaderno, no. 87, pp. 1–32 (2012)
8. Quattrone, G.: Protocollo definizione e mitigazione dei livelli di rischio dei beni, report della
ricerca CNR-IIA (WP7) per il progetto SIMONA SIstemi e tecnologie per il MONitoraggio
di Aree culturali in ambiente subacqueo e terrestre, POR Calabria FESR 2007/2013 (2015)
9. Quattrone, G.: Planning religious tourist routes for the development of Calabria territory. In:
Trono, A. (ed.) Sustainable Religious Tourism. Esperidi Ed. Lecce (2012)
10. Taboroff, J.: Cultural heritage and natural disasters: incentives for risk management and
mitigation. In: Kreimer, A., Arnold, M. (eds.) Managing Disaster Risk in Emerging
Economies, the World Bank, Washington, pp. 233–240 (2000)
11. Teti, M.A.: La pianificazione delle aree archeologiche. Carta dei vincoli archeologici della
Calabria (1912–1992). Gangemi editore, Roma (1993)
12. Will, T., Meier, H.R.: Cultural Heritage and Natural Disasters: Risk Preparedness and the
Limits of Prevention ICOMOS. Heritage At Risk, Dresden (2007). Special Edition
Full Paper: Intangible Cultural Heritage
Documentation
Parameterizing the Geometry and Visualizing
the Lighting Method of Byzantine
Church Domes
Wassim Jabi1(&) and Iakovos Potamianos2

1
Welsh School of Architecture, Cardiff University, Cardiff, UK
[email protected]
2
Faculty of Fine Arts, School of Drama,
Aristotle University of Thessaloniki, Thessaloniki, Greece
[email protected]
Abstract. This paper introduces a computer-based tool for the analysis of the
geometry and the daylighting of Byzantine church domes to facilitate experi-
mentation with a number of cases before any fieldwork is undertaken. Starting
with a geometric derivation of the relationship between dome parameters, the
digital tool builds an interactive three-dimensional model of a Byzantine church
dome. The model allows the user to input the properties of the dome, the drum,
any windows, and the slope of their sills. The model allows the user to define the
dome using three different curvatures since such a case was identified in a
Mistras church. A custom ray-tracing algorithm visualizes the path of light rays
falling on the windowsills and their reflections within the dome. It was found
that several parameters are interrelated and that an optimal set of proportions
must be established to achieve the expected behavior of light within the dome.
Keywords: Byzantine domes Lighting simulation Parametric models

Natural daylighting
1 Introduction
In many Byzantine churches the dome appears to be uniformly illuminated even though
logically it could not be since the windows are located at the lower part of the dome and
the sunrays always point downwards (Fig. 1). According to our studies this phe-
nomenon was first posed and solved as a problem of solid geometry by Anthemius of
Tralles presumably to achieve the unique lighting of the Hagia Sophia dome, his life
masterpiece. Anthemius was an important geometer who was well versed in “catop-
trics”, a science, which captured the interest of mathematicians since classical antiquity.
While the problem was clearly posed by Anthemius, there is no record that the com-
bination of forms he invented was known or followed in subsequent churches. How-
ever, although later churches adopted a quite different geometry, the phenomenon of a
highly lit dome persisted. In order to understand the manner in which a byzantine dome
was lit, we have attempted several studies and simulations [1–5], of which the one
appearing in this paper constitutes a new development.

DOI: 10.1007/978-3-319-48496-9_14
172 W. Jabi and I. Potamianos
Fig. 1. Dome of the Iveron monastery church, Mount Athos, 11th c.
The dome of Hagia Sophia of Istanbul, as has been established in our previous
studies, had been designed as a combination of two curved mirrors (Fig. 2) of different
curvature [1, 2]. These mirror designs were found in the architect’s own extant writings
[6]. The manner in which these two mirrors were combined has been shown in previous
studies (Fig. 3), and the impression generated has been described by Procopius, the
historian eye-witness of the effect. This hypothesis can be considered largely proven in
terms of geometric, topographic and textual evidence [1].
Fig. 2. a. Anthemius’s ellipsoidal reflector, b. Anthemius’s spherical reflector (i.e. the shallow
dome itself), c. Sketch showing the location of the reflectors.
The most persistent questions posed, whenever these findings had been presented in
academic fora, were the following: “Hagia Sophia is a unique case. How would domes
of lesser church examples achieve a similar visual effect? Since their form is quite
different than Hagia Sophia’s and thereby could not have employed the same method of
illumination, how did they achieve the constantly and uniformly luminous effect which
is common among them?”
Parameterizing the Geometry and Visualizing the Lighting Method 173
Fig. 3. Computer model of the original dome of Hagia Sophia with the two mirrors in place
(from an earlier simulation).
2 The Case of Other Byzantine Churches
Any attempt to answer the above questions tends to become particularly complex. The
reason is that one is not dealing with a single building, which may be studied
exhaustively from a number of different aspects. Instead, one has to investigate many
different buildings of various eras. Therefore, the enormity of the problem requires a
carefully planned strategy.
A highly and quite uniformly lit dome in lesser Byzantine church examples is a
persistent phenomenon. This is not something expected or occurring in all domed
churches. If there had been no lighting planning, a different portion of the dome would
be lit depending on the position of the sun in the sky (Fig. 4) while the rest of the dome
would remain darker. Instead, a Byzantine dome is uniformly lit and the brightest spot
is usually found at the apex (Fig. 1). Therefore, it may be assumed that there must be a
certain lighting system at work in the domes of lesser Byzantine churches as well. This
system is what we set out to discover and to this purpose we thought that we could take
advantage of the method of parametric modeling before we move into actual field
work.
Fig. 4. Drawing of the original Hagia Sophia dome by E. Antoniades, showing a bright spot in
the dome, located off center [8]. This image would be created if the lighting had not been
carefully planned.
Following the construction of the parametric model one could then resort to surveys
of existing domes. In typical surveys of Byzantine churches, published in various
specialized articles, books, and dissertations, some basic, easily measurable variables
are already available. Thus, one would have a significant amount of elementary
measurements, (i.e. dome height, drum height, number and dimensions of dome
windows) to begin with, before travelling to the monument site. One would then have
to observe whether the dome of the monument is luminous, using existing photographs,
and would manipulate the subtler variables of the parametric model in order to achieve
the effect. Once the luminous effect had been achieved in the model one would already
have a fair estimate of the variables that affect this visual impression. As a result, only
those variables of the actual form would have to be measured that made sense in terms
of the model. Strategically, this would be the most effective procedure.
3 Geometric Principles of the Parametric Model
The Byzantine church domes vary considerably. They are made up of various parts.
Often, there is an upright standing tube-like part, which may be cylindrical or polyg-
onal, called the “drum”. On top of it a roughly semispherical dome is placed. On the
drum several openings are pierced which differ in number from six to eight to twelve to
forty. However, the closer one looks at that geometry the more variations one discovers
to this basic scheme.
Below, certain general observations and principles are outlined which lie behind the
construction of the model. First, the drum is not always of the same height and its
proportions are related to the scale of the dome ensemble (i.e., the diameter of the dome
in plan) as well as to the total height. A general proportioning rule is that a greater
dome radius corresponds to a lower drum height. This rule has not yet been established
as a strict mathematical proportional relationship.
In some cases, the drum disappears completely. This is the case, for instance, of the
original dome of Hagia Sophia in Istanbul (Fig. 3), while in the church of Saints
Sergius & Bacchus – which served as a model for the design of the former - the drum is
quite low. In the church of Hagia Eirini, of the same city, while a substantial drum is
visible from the outside, in the interior the smooth curvature of the dome blends in with
the drum and makes the drum invisible as if it were part of the dome. This difference is
visible in most churches of later periods as well. While the drum is clearly articulated as
a form from the exterior, it loses all articulation and blends in with the curvature of the
dome as it is seen from the inside. Architecturally, the exterior and the interior seem as
belonging to two different aesthetic approaches. This phenomenon requires a special
investigation, which, however, is not the focus of this paper. What we might retain
from this observation for now is that for some reason this sense of transition from one
geometric form to another, so that the composite form appears unified, as seen from the
inside, constituted an essential goal.
Second, the size of the windows piercing the drum depend on the drum’s height.
The taller the drum the greater the height of the windows and the slenderer their
proportions become. This proportional relationship appears to be consistent. However,
as we have already mentioned, the drum in some cases may not exist at all. In these
cases, the window height acquires a proportional relationship to the sum of the circular
segments length. Therefore, the possibility of inserting windows on the dome curvature
should be incorporated in the model. The window height does have an effect on the
manner in which light is reflected towards the dome. In a taller drum the rays seem to
be able to be reflected more vertically while in a shorter drum the window header limits
the reflected rays to more oblique angles.
Third, it has been noticed that the dome may not be necessarily semispherical.
What has been observed in Hagia Sophia is that a very shallow curvature was
employed. Such shallow curvatures have been observed in various other churches as
well. Similar shallow curvatures were not limited to the dome but were employed also
in vertical wall surfaces such as the one above the main gate of the church of Hagia
Sophia, which seems to be responsible for the concentration of light so that the image
of Christ, located on that surface, becomes suddenly lit at a certain distance [7].
Shallow curvatures, in general, seem to be connected to the impression of a uniformly
lit dome. A shallow curvature has the ability to trap light and distribute it uniformly
onto its surface, as long as light enters the curvature obliquely. This particular char-
acteristic would render the shallow curvature invaluable to a dome designer. Therefore,
our parametric model has the capability of incorporating a shallow dome curvature.
Fourth, a sunray passing through a dome window must somehow be reflected in
order to reach the interior of the dome. The windowsill may provide such a surface.
Anthemius especially designed such a surface for this location [1, 4]. However, this
design, placed on the dome windowsill was conceived as a concave surface, which
sloped inward. For the case of Hagia Sophia, of which the original dome had no drum,
this windowsill design and slope was reasonable. For a dome with a drum, however,
the same principle would tend to produce an outward sloping windowsill. We have
found evidence of such treatment on the dome windowsills of the Monastery of Iveron
church, which has a particularly bright and uniformly lit dome (Fig. 1). Based on these
observations, the model has the capability of manipulating the windowsills by changing
their slope and curvature.
Fifth, if not planned carefully, the reflected ray, once it reached the dome, would
most probably hit it once and then be reflected downward to the floor. The point on
which it hit would vary depending on the position of the sun in the sky. This means that
a luminous point in the dome would appear at different positions while the rest of the
dome would remain dim (Fig. 4). It appears that the only way to create a uniformly and
constantly luminous dome would be to trap the light in it. This can happen if two
conditions are met. First, if a shallow reflective surface for the interior of the dome
itself is employed. Second, if a windowsill with a slope is employed so as to send the
rays into the dome in such a way that further reflections continue to remain entrapped
within the dome. Then, we could expect that most ordinary churches may incorporate
not one but two or even three curvatures within the dome. The upper curvature of the
dome may have to be deeper when a drum exists and be proportional to its height. The
idea of a tri-partite dome was conceived based on the observation of an existing dome,
of the Metropolis church of Mistras (Fig. 5).
Fig. 5. Dome of Metropolis church, Mistras, 13th c.
3.1 Derivation of the Geometric Relationships

The purpose of this section is to provide a geometric proof of the derivation of the two
other radii (called middle dome radius and upper dome radius) given the first radius of
the lower/main dome and a set of input parameters. This derivation forms the basis of
the algorithms and the parametric calculations based on user input in the interactive
graphical interface. Please refer to the diagram below for reference to the named
vertices (Fig. 6).
Fig. 6. Drawing showing the section of the dome which is composite of three circular
curvatures. This drawing shows the fundamental variables of the dome introduced into the model.
Before we start the derivation, it is important to note that the naming of the domes
(lower, middle, and upper), refers to the end result after its construction. These domes
are shown in dashed lines in the figure above. The actual domes need to be vertically
stacked in the opposite direction to achieve that end result (shown in a solid line). The
lower dome arc has the smallest radius (r1) and the highest vertical position for its
center (A), followed by the middle dome with a larger radius (r2) and a lower vertical
position of its center (B), and ending with the upper dome with the largest radius (r3)
and the lowest vertical position of its center (C).
We will mainly use Pythagoras’s theorem which defines the relation in Euclidean
geometry among the three sides of a right triangle. It states that the square of the
hypotenuse (the side opposite the right angle) is equal to the sum of the squares of the
other two sides. In addition, we will use trigonometric functions to derive the angles
subtended by the arcs we are calculating. The calculated lengths and angles will allow
us to geometrically create the necessary arcs of any desired dimension while main-
taining consistent parametric relationships in the overall geometry.
Given parameters
The derivation assumes a set of given parameters:
1. r1 is the radius of the lower dome. This is the main interior radius of the dome as
perceived from the bottom at the “springing” point of the dome.
2. AB is the vertical distance from the centre of the lower dome to the centre of the
middle dome
3. BC is the vertical distance from the centre of the middle dome to the centre of the
upper dome.
4. ED is the vertical height of the lower dome portion. Above that height, the dome
arcs according to the middle dome radius (r2).
5. HG is the vertical height of the middle dome portion. Above that height, the dome
arcs according to the upper dome radius (r3).
Derivation of the radius of the middle dome (r2)
r12 ¼ AE 2 þ ED2 ð1Þ
Since ED and r1 are given, we can derive AE.

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
AE ¼ ðr12 ED2 Þ ð2Þ
BF ¼ AE ð3Þ
AE is calculated in (2)
FD ¼ FE þ ED ð4Þ
Both FE (AB) and ED are given, so we can calculate FD.
r22 ¼ BF 2 þ FD2 ð5Þ
Since BF and FD have been calculated above in (3) and (4), we can derive r2.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
r2 ¼ ðBF 2 FD2 Þ ð6Þ
Derivation of the radius of the upper dome (r3)
JG ¼ JI þ IH þ HG ð7Þ
JI ¼ AB; IH ¼ ED ) JG ¼ AB þ ED þ HG ð8Þ
AB, ED, and HG are given therefore we can derive JG.

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
r22 ¼ JG2 þ BJ 2 ) BJ ¼ ðr22 JG2 Þ ð9Þ
Since r2 and JG have been calculated above, we can derive BJ.
CK ¼ BJ ð10Þ
BJ is calculated in (9).
KG ¼ KJ þ JI þ IH þ HG ð11Þ
KJ ¼ BC; JI ¼ AB; IH ¼ ED ) KG ¼ BC þ AB þ ED þ HG ð12Þ
BC, AB, ED, and HG are given, therefore we can derive KG.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
r32 ¼ CK 2 þ KG2 ) r3 ¼ ðCK 2 þ KG2 Þ ð13Þ
CK and KG have been calculated in (10) and (11) respectively, therefore we can
derive r3.
Thus, given the radius of the lower dome, the vertical position of the centres of the
three domes and the vertical height of the lower and the middle dome portions, we can
derive the radii of the lower and higher domes and construct a complex dome curvature
made out of three arcs that can have a significantly lower apex (N) compared to that of
a single-radius dome (L), or a double-radius dome (M).
Derivation of the angles
The construction of the curves of the parametric dome also depend on discovering the
subtended angles for the lower, middle, and upper dome. In the figure above, these
angles are <EAD (α), <DBG (β), and <GCN (γ) respectively. Using trigonometry, we
can find these angles. Specifically, we can use the definition of the sin of an angle as the
division of the opposite side divided by the hypotenuse in a right-angled triangle. Space
limitations prevent us from including the full derivation in this paper.
4 Computer Implementation
In order to investigate these hypotheses, we created a computer script within Autodesk

3 ds Max 3D that generates a three-dimensional solid model of a dome (Fig. 7).
The parametric model presented here has been constructed with the capability of
representing a dome created with three different arcs above the drum. The dome can
Fig. 7. Parametric dome model and associated input parameters.
acquire any radius and wall thickness. The lower and upper curvatures of the dome can
vary in terms of their vertical height; thus creating different composite curvatures while
maintaining a consistent geometric relationship between them. Any number of win-
dows, of any width and height, can be incorporated into the drum or into the lower
curvature of the dome. The windowsill can be made to slope inward or outward.
Once the dome has been created, we created a custom ray-tracing algorithm to
simulate the behavior of light and reflection to visualize a representation of the sunrays
as they hit the windowsill and reflect multiple times inside the dome (Fig. 8). This
methodology is similar to the ones used for acoustic analysis [9]. The accuracy of the
computer modelling has previously been verified [10]. In the image below, two
sub-cases are presented simultaneously, one with a semispherical upper dome and
another with a shallower upper dome in order to compare the manner in which light is
reflected in each. The light is reflected in such a way that it washes the entire dome
surface. This light behavior guarantees the uniform illumination of the entire dome.
Fig. 8. Computer model comparing a semispherical and a tri-spherical dome
We decided that the most effective method to study the behavior of the multiple
reflections within the dome was to calculate and visualize the line vectors that represent
the direction of a sunray. It is important to note here that the built-in ray-tracing
capabilities of the software environment were of no use to us at this stage as they only
represent the final brightness and color of surfaces, but not the light ray vectors
themselves which we needed. Furthermore, we also abandoned the use of the built-in
physics simulation engine as reported in a previous paper. The reason for developing a
ray-tracing algorithm rather than rely on the built-in physics engine is to eliminate data
noise present in the physics simulation that takes into consideration the friction of
surfaces, bounciness and other forces. In light-simulation, these forces do not exist.
Thus, the enhanced script uses pure 3D vector geometry and rules of reflection to
calculate the path of light as it reflects multiple times within the dome. While outside
the scope of this paper, once the geometry is derived based on the mathematics of light,
additional visual studies of texture and spatial perceptions can be undertaken following
the examples in the project “Virtual Hagia Sophia” at MIRALab [11]. Similarly,
high-dynamic-range (HDR) imaging can be used as a good method to present a similar
range of luminance to that experienced through the human eye and has been used to
better visualize light beams in existing buildings [12].
By varying the input parameters and re-running the ray-tracing algorithm, we were
able to simulate multiple incidence angles and derive the appropriate parameters that
can achieve the desired light path and therefore luminous effect through the entrapment
of light within the dome (Fig. 9).
Fig. 9. Computer model of a dome with drum and a convex parabolic windowsill.
In addition to the parametric script that creates the main geometry of the dome, we
enhanced our work with an additional script that allows us to generate and test the
effect of parabolic surfaces. The script accurately creates a parabolic surface based on
the location of a “focus” point and a “directrix” line. This three-dimensional parabolic
surface can then be placed on top of the sill window and incorporated in the dome
geometry. The axis of the parabola (i.e. its general direction) can be oriented in any
angle. The orientation of the main axis of symmetry of the parabola determines the
direction of reflections for any ray that falls on the exterior surface of the parabola and
passes through its focus if extended to the other side of the parabola. However, parallel
rays that hit the surface in other areas will reflect in different and non-parallel directions
because they do not pass through the focus if extended. Using these scripts, we were
able to simulate and visualize the behavior of several parallel light rays coming in at a
steep angle as can be seen in (Fig. 9). The two rays in the above figure, the first exiting
downwards on the right side (indicated in red) and the second exiting the window on
the right side (indicated in blue), are close to each other and thus behave in a similar
fashion. These two rays enter the dome space at a steeply vertical angle which causes
them to reflect along the surface of the dome in a tangential matter. Other parallel rays
that do not pass through the focus of the parabola are reflected in other directions —
mainly back to the exterior of the dome. With the introduction of these additional
scripts and the refined ray-tracing algorithm, we now have more accurate tools to
further study different reflecting geometries and propose possible methods used to light
the apex of the dome.
5 Summary of Findings
So far we have established the following:

1. In order to generate a light wash of the dome surface the rays must enter the
concave surface as laterally to the dome curvature as possible. In a dome with a
drum, this lateral introduction of the reflected ray corresponds to the vertical
direction.
2. The taller the drum the more possible it is for the reflected ray to enter the dome at
a more vertical angle.
3. The variable that is most important for the upward lateral direction of the reflected
sunray is the slope of the windowsill of the dome windows. In a dome with a drum
the windowsill must slope outwards while in the absence of a drum it must slope
inwards.
4. In order for the reflected sunrays to acquire the same upward direction regardless of
the position of the sun in the sky, the windowsill must be curved and most probably
convex, acquiring the form of a parabola.
5. Both a semispherical dome as well as a tri-spherical one will generate consecutive
reflections within it, given the appropriate windowsill shape.
6. In a tri-spherical dome the transition from one curvature to the next must be as
gradual as possible in order for the reflections to follow the dome curvature closely.
7. The difference in a tri-spherical dome appears to be that more reflections are
concentrated toward the upper parts of the dome.
8. Various parameters are interrelated. The height of the drum to the number and
radius of the curvatures of the dome as well as to the rotation of the central axis of
the convex parabolic windowsill.
9. Every sunray that falls in the direction of the focus of the parabola is reflected
upwards within the dome and begins travelling along its curvature.
10. The closer the first reflection is to the vertical the greater the entrapment of the
reflections within the dome.
11. Most of the sunrays that do not align with the parabola focus are reflected back out.
Therefore, this type of curvature creates a highly selective and exclusive
windowsill.
6 Conclusions and Future Work
Byzantine church domes are paradoxically brighter than all surrounding surfaces. This
phenomenon has been investigated in previous work for the case of the dome of Hagia
Sophia of Istanbul. The lighting of this dome had been resolved by the combination of
two reflectors designed by Anthemius. In order to understand the dome design methods
employed in lesser church examples, where a luminous dome effect is achieved, a
parametric model was constructed which gave us the opportunity to experiment with a
number of parameters influencing the behavior of light within the dome before we test
any particular church case. It has been found that several of the parameters are inter-
related, as for instance, the height of the drum, the slope of the windowsill and the
height of the three curvatures of the dome. Depending on the various forms constituting
the dome an optimal set of proportions must be established for the behavior of light
within the dome. It is probable that there is a precise mathematical relationship gov-
erning these factors. This is an area to be investigated in future work. Such an
investigation is valuable along with archaeological evidence to the restoration of domes
of Byzantine churches of various periods. Contingent on the availability of funding, we
plan to use advanced laser scanning and photogrammetry techniques on site to docu-
ment the geometry of selected domes and their windowsills to verify our parametric
model findings. Given the height of windowsills, we anticipate the need for
aerial-based photogrammetry (e.g. using quadcopter drones equipped with cameras and
3D laser scanning equipment). Ultimately, we would like to study the experiential
qualities achieved through the manipulation of light and its interaction with the
architecture.
References
1. Potamianos, I.: Light into architecture: the evocative use of natural light as related to liturgy
in Byzantine Churches. Ph.D. dissertation, Doctoral Program in Architecture, University of
Michigan, Ann Arbor (1996)
2. Jabi, W., Potamianos, I.: A parametric exploration of the lighting method of the Hagia
Sophia Dome. In: Proceedings of 2006 7th International Conference on Virtual Reality,
Archaeology and Intelligent Cultural Heritage, pp. 257–265 (2006)
3. Jabi, W., Potamianos, I.: Geometry, light, and cosmology in the church of Hagia Sophia. Int.
J. Architectural Comput. 5(2), 303–319 (2007)
4. Potamianos, I., Jabi, W.: A parametric model of Byzantine church domes. In: Proceedings of
the 6th Balkan Light Conference, pp. 13–17 (2015)
5. Jabi, W., Potamianos, I.: Interactive parametric design and the role of light in Byzantine
Churches. In: Proceedings of 24th eCAADe Conference, pp. 798–803 (2006)
6. Huxley, G.L.: Anthemius of Tralles: A Study of Later Greek Geometry. Eaton Press,
Cambridge (1959)
7. Mathew, G.: Byzantine Aesthetics. Viking Press, New York (1964)
8. Antoniades, E.: Ekphrasis tes Hagias Sophias (1905)
9. Turner, J., Hall, T.W.: An application of geometric modeling and ray tracing to the visual
and acoustical analysis of a municipal open-air auditorium. In: Proceedings of ACADIA
1990, pp. 173–185 (1990)
10. Taengchum, T., Chirarattananon, S.: Ray tracing method of light through rectangular light
pipe with bends. Energy Procedia 79(2), 791–798 (2015)
11. Foni, A., Papagiannakis, G., Cadi-Yazli, N., Magnenat Thalmann, N.: Time-dependant
illumination and animation of virtual Hagia-Sophia. Int. J. Architectural Comput. (IJAC) (IF:
1.144) 5(2), 284–301 (2007). Multi-Science Publishing
12. Happa, J., Artusi, A., Czanner, S., Chalmers, A.: High dynamic range video for cultural
heritage documentation and experimental Archaeology. In: Proceedings of 2010 VAST 2010
the 11th International Symposium on Virtual Reality, Archaeology and Cultural Heritage
(2010)
Digital and Handcrafting Processes Applied
to Sound-Studies of Archaeological Bone Flutes
Etienne Safa1 , Jean-Baptiste Barreau2,3 , Ronan Gaugne4 ,

Wandrille Duchemin6 , Jean-Daniel Talma7 , Bruno Arnaldi3 ,
Georges Dumont5 , and Valérie Gouranton3(B)
1
Université de Bourgogne/ArTeHiS UMR 6298, Dijon, France
2
CNRS/CReAAH UMR 6566, Rennes, France
3
INSA de Rennes/IRISA UMR 6074/Inria-Rennes, Rennes, France
[email protected]
4
Université de Rennes 1/IRISA UMR 6074/Inria-Rennes, Rennes, France
5
ENS de Rennes/IRISA UMR 6074/Inria-Rennes, Rennes, France
6
LBBE UMR CNRS 5558, University of Lyon 1, Lyon, France
7
Atelier El Bock, Chatel-Montagne, France
Abstract. Bone flutes make use of a naturally hollow raw-material. As

nature does not produce duplicates, each bone has its own inner cavity,
and thus its own sound-potential. This morphological variation implies
acoustical specificities, thus making it impossible to handcraft a true
and exact sound-replica in another bone. This phenomenon has been
observed in a handcrafting context and has led us to conduct two series
of experiments (the first-one using handcrafting process, the second-one
using 3D process) in order to investigate its exact influence on acoustics
as well as on sound-interpretation based on replicas. The comparison of
the results has shed light upon epistemological and methodological issues
that have yet to be fully understood.
This work contributes to assessing the application of digitization, 3D
printing and handcrafting to flute-like sound instruments studied in the
field of archaeomusicology.
Keywords: Acoustics · Statistics · Handcrafting · Raw-materials ·

Digitization · 3D printing · Music archaeology
1 Introduction
Elaborating a research project in close collaboration with a craftsman and a
research team dedicated to digitization of cultural heritage was the trigger point
to different kinds of experiments meant to investigate the morphological vari-
ability of bones and its influence on the emitted sounds when carved as flutes.
Dealing with this “Sound-morphology” is the main part of a craftsman’s work,
which is why it was decided to run the project of an apprenticeship that would
last for one year [18]. During this time, particular attention was paid to the
creation and use of prototypes, i.e. a bone flute manufactured in order to try

DOI: 10.1007/978-3-319-48496-9 15
Digital and Handcrafting Processes Applied to Sound-Studies 185
and understand the sound specificities of a particular bone, and then used as a
guide in order to ease the adaptation process. Indeed, each bone has its own mor-
phology and needs to be considered as an individual. The flute-maker proceeds
then with a precise observation of each individual and takes every morphological
specificity into consideration in order to craft series of bone flutes with similar
sounds and identical tuning, even if this has to result in objects that do not look
the same. Otherwise, he would risk to create an inefficient object, or a completely
different flute.
These observations have raised specific issues regarding the use of bone flute’s
replicas for tone scales interpretations in archaeological surveys, as their manu-
facture never seems to take into consideration the bone’s morphology as part of
its acoustical specificities [6,8,16]. They have also led us to conduct “twin exper-
iments” in the hope of reaching consistent results that would spare no method-
ological tracks (past, actual and yet-to-come sound-reconstruction methods) in
order to explore their limitations as well as their potential. This way, we hope
to contribute to better the epistemological landscape of archaeological flute’s
research.
The work presented in this paper focuses on the comparison of the sound
results given by both series of experiments.
2 Context of the Work
2.1 Approach
Flutes are not all the same. They are grouped into several kinds which are distin-
guished by the way the air stream is directed toward the edge. Each kind has its
own sound aesthetics, but gives also more or less freedom to the flute-player in
choosing the pitch and the sound’s characteristics, thanks to the blowing angle
variability (Fig. 1). Oblique-, pan-, vessel- and transverse-flutes are amongst the
most malleable kinds of flutes. We chose duct-flutes as they are the opposite.
In term of organology, these objects can be mentioned as 421.221.12 in the
S/H classification system (Sachs/Hornbostel), which means: Internal duct-flute
(straight and single) with finger holes and an open end.
Fig. 1. Blowing angle variations regarding two different organological kinds of flutes:
(a) oblique flute, (b) duct-flute
186 E. Safa et al.
2.2 Partnership
This “two-front approach” demands to assemble a consistent amount of knowl-
edge, which can only be achieved through partnership.
• Handcrafting process: the work gathered a traditional flute-maker and a
statistician in computational biology.
• 3D process: the work was based on an existing collaboration between ar-
chaeologists and computer scientists on advanced imaging for archaeology,
the CNPAO [2]
2.3 Terminology
This paper will use the following terminology according to the acoustical speci-
ficities of bone flutes:
• Morphology: refers to the natural inner and outer shapes of the bone.
• Geometry: refers to the handcrafted inner and outer shapes carved deliber-
ately or not onto the bone’s surface.
• Sound-morphology: refers to the acoustical sections of the morphology,
which define the sound potential of the bone (i.e. the inner cavity). By defin-
ition, each bone has a different sound-morphology.
• Sound-geometry: refers to the acoustical sections of the geometry, which
are involved in the definition of the instrument’s final sound, whether they
were meant (deliberately carved) or not (unintentional and/or unconscious
geometry). As an example: the shapes of the internal duct, of the edge, of
the finger holes, etc. By definition, the sound-geometry rules out the outer
shaping as long as it does not change the finger holes depth.
• T0, T1, T2, etc.: refers to the finger holes’ combination. T0 means all holes
closed. T1 means that the lower finger hole (the first one) is open. T2 means
that the two lower finger holes (the first and the second one) are open, etc.
• F0, F1, F2, etc.: systematic identification numbers of the experimental
flutes. F0 refers to the control flute, whereas F1, F2, F3, etc. refers to each
replica copying the control flute.
2.4 Related Works

Nowadays, 3D technologies allow outer and inner contact-free investigation on
complex geometries [15]. As such they contribute to answer both preservation
and sound studies issues and are more and more used in the actual archaeo-
musicological research. If their consequences on our interpretations are still to
be defined, they allow different kind of approaches and studies that aim to get a
better understanding of ancient sounds. They can be applied to any organological
material [10], such as string instruments [4,13,20,21] but also aerophones [3,8,
9,11], among which archaeological “flutes”, and objects presumed to be flutes,
figure [1,14,22,23].
Eventually, the music-archaeology research may even explore new possibilities
in sound reconstruction studies, as its data can be applied to sound simulators
and sound-scape reconstructions [7,12,24].
3 The Sound-Morphology Principle

Naturally hollow rawmaterials, such as bones, hornes, shells or reeds, present a
morphological variability between one individual and another. Those variations
can be observed both regarding their shapes, their scale and their volumet-
ric and spatial configuration (Fig. 2). Some of them are involved in the sound-
morphology. For example, a larger bone will produce a lower pitch for the same
length. Likewise, an important and sudden increase or decrease of the bone’s
conicity tends to distort the efficiency of a close-range finger-hole.
Fig. 2. Morphological variations between bones used for crafting F0 to F5 in the Hand-
crafting process experiments explained below. Deer femurs show several constants, such
as a bulge characterizing the distal part of the epiphysis, a triangular and irregular
depression characterizing its proximal part, and a slimming zone in the concave area of
the bone’s bean-like cross-section. Despite those constants, there never are two identical
bones.
In order to illustrate this phenomenon, we chose to handcraft a unique replica

of a bone flute in another similar bone (Fig. 3).
The control flute was made in a goat’s tibia. It was made very simply, using
only steel knife and file, evoking archaeological flutes found in northern Europe
for medieval period [5]. The handmade replica was made very carefully, using
several measurement tools (caliper, compass, etc.). Also, as the depth of the
block changes the pitch, we chose depth 0 (Fig. 4). This calibration is easier to
reproduce. We also tried our best to give both blocks a similar soil angle. As
a result, the two flutes gave different sounds, with a deviation going from half
a tone to more than one tone, increasing as we open the finger holes (Figs. 9
and 10).
This replication test shows how much the sound of a bone flute replica may
be deviant from the sound of the control flute it’s related to. This phenomenon
illustrates the notion of “sound-morphology” as it reveals that every bone has a
sound-potential of its own.
188 E. Safa et al.
Fig. 3. The control flute (left) and its replica (right) both made out of goat’s tibias.
4 Handcrafting Experiments
4.1 Handcrafting Replication Process and Technical Specifications
• Objectives: those experiments aim to define the extent of the limitation
caused by sound-morphology, as well as to explore the acoustical specificities
of this phenomenon. The approach is then different from what we can see
in experimental archaeology, as we need here a well-known, functional and
replicable bone flute in order to compare its actual sounds with our interpre-
tations.
• Control and sample: we chose 6 similar deer femurs with morphological
variations. 5 replicas is the minimum sample required for statistical analysis.
• Chosen sound-geometry: inner duct-flute with rectangular opening and
straight edge (Fig. 5). Combined with a straight geometry, this configuration
creates powerful blowing constraints and is easier to reproduce.
• Manufacture: handcrafted in January 2016.
• Sound capture and analysis: because of lack of means, we had to use a
common recording device (smartphone) and a free software (audacity). Having
no mechanical blower nor anechoic chamber available at the time, we had to
record the sound using natural blowing (as homogeneous as possible) and
the same context (a chosen room). Thankfully, the studied phenomenons are
contrasting enough to be well illustrated even with a lack of technical means.
Fig. 4. Illustration of Depth 0 and soil Fig. 5. Depiction of the sound-

angle geometry used for the handcrafting
experiments
4.2 Sound Results

The diagrams in Fig. 8 represent the results of basic acoustical analysis of the
control flute and its 5 replicas. They obviously show that each individual is
different from the control flute.
4.3 Statistics and Discussion

The table (Table 1) represents statistical analysis made on the recorded frequen-
cies. In order to compare them properly, we had to translate them from Hertz
to logarithmic scale (base 2 logarithm).
This table shows heterogeneous frequencies and intervals deviations compar-
ing the sample to the control flute, as well as between each individual from the
sample itself. Even if the frequency deviations are mostly non-significant regard-
ing statistics (T0 is the only one being significant), the sound estimation they
produce is not satisfying for the ear (about one quarter-tone). However, inter-
vals deviations are really small in comparison (about 1/20th of a tone), which is
extremely accurate.
The following facts should also be considered regarding those results:
1. The lower end of the flute was one of the most variable areas and it was then
difficult to reproduce an exact geometry in a changing trabecular bone. This
could explain T0 deviation.
2. The small sample size is probably involved in those statistical results: a larger
sample (20 to 30 replicas) should help us to get better results and thus assess
if whether or not this incredibly accurate estimation of intervals is exact. It
should also explain the difference between a satisfying intervals reproduction
and an unsatisfying frequencies reproduction.
3. The human blow should be ruled out and replaced by a mechanical blower in
order to ensure the accuracy of the sound-capture.
190 E. Safa et al.
Table 1. Statistical analysis of frequencies emitted by F0 to F5 while playing succes-

sively T0, T1 and T2. Differences are expressed in semi-tones (“−1” equals “1 semi-tone
lower”). The right columns show intervals deviations (T0–T1 and T1–T2).
5 3D Experiments
5.1 CT-scanning
There exist several possibilities in matter of 3D image acquisition, but CT-
scanning was the only viable option because of the very nature of flutes: inner
shapes are drastically important and their acoustical properties are extremely
sensitive. We needed then a technology that would be able to capture high
resolution images both inside and outside of the objects. µ-tomography, also
known as µCT, was then the perfect tool. This technology uses X-rays in order
to recreate high resolution 3D internal views of an object by compiling the
acquired images and is mainly used in medical imaging and industries.
5.2 3D Replication Process and Technical Specifications
• Objectives: those experiments aim to question the sound-replication capa-

bility of 3D technologies in order to define whether or not they may allow
us to pass beyond the sound-morphology limitation endured by handcraft-
ing process. They also aim to assess their own limitations and potential as a
sound-reconstruction method.
• Technologies used:
1. µCT-scanning: the machine is an X-ray microfocus CT system Gen-
eral Electric (formerly Phoenix) v—tome—x 240D from CRT Morlaix, a
resources center dedicated to metrology (http://www.crt-morlaix.com/).
In the set-up, the sample is placed on a rotating table, and the X-ray
source and detector are stationary.
2. 3D wire and resin printing: the machines are a MakerBotReplicator2
from IUT Le Creusot, and a Stratasys Mojo from ENS Rennes. The resin
model was printed on a 3D Objet by a contractor.
• Scanned object: we chose to scan the control flute used in the sound-
morphology principle (the one made from a goat’s tibia) in order to compare
the 3D results to the handmade replica. The flute was scanned in three parts
in order to get a precision of less than 50µ. The reassembly was processed
Fig. 6. Disconnected objects (yellow) Fig. 7. 3D sculpted patch (transpar-

in the area of the trabecular bone. ent gray) on Blender (based on the
(Color figure online) geometry of the cloud).
with the software Autodesk Meshmixer. Also, as the trabecular bone renders
through µ-CT scanning as a cloud of 600+ tiny objects, it cannot be directly
printed (Fig. 6). We chose to explore two possibilities: simply removing the
objects in one case, and integrating them as a 3D sculpted “patch” in the
other (Fig. 7). We used Meshlab and Blender in order to get ready-to-print
3D models.
• Replicas: F1 refers to the handmade replica. F2 refers to the 3D orange wire
replica (with 3D sculpted “patch”, no post-printing treatments). F3 refers to
the 3D white wire replica (without the trabecular bone, acetone bath and
ultrasounds post-printing treatment). F4 refers to the 3D white resin replica
(better printing resolution, with 3D sculpted “patch”, no post-printing treat-
ments).
• Printings: printed between January and May 2016.
• Sound capture and analysis: same context than for the handcrafting
process.
5.3 Sound Results

The diagrams in Fig. 9 represent the results of basic acoustical analysis of the
control flute and its four replicas.
5.4 Analysis and Discussion
The following tables represent sound-comparisons between the control flute and
its replicas using the recorded frequencies translated from Hertz to base 2 loga-
rithm (Table 2).
As we expected, this table shows that 3D printed replicas are globally closest
to the original than the handmade one. This is due to the absence of the bone’s
morphological variability that would occur from using several bones. However,
they are not identical between each other (Fig. 10).
192 E. Safa et al.
Fig. 8. Diagrams analysis of F0 to F5 Fig. 9. Diagrams analysis of F0 to

F4
Table 2. Comparison between frequencies (top)/intervals (bottom) emitted by F0 to

F4 while playing successively T0, T1, T2, T3 and T4. Green cells indicate a sound-
reproduction precision of 1/20th of a tone or less.
In Fig. 10, both orange and white wire flutes present a significant but different
deviation regarding their emitted frequencies, whereas the resin flute is the most
accurate of them all. Indeed, it reaches the sounds of the original with a precision
of less than 1/20th of a tone.
As it appears, acoustical phenomenons related to 3D printed replicas seem
to be quite intricate. The following facts should thus be considered regarding
those results:
1. 3D wire-printing is processed by fusing a plastic filament which is then

deposited by layers, and finally cools down and solidifies. The cooling process
comes with a shrinking phenomenon which extent depends on the wire itself
as well as on the cooling context (hygrometry and temperature) [17]. Fur-
thermore, these deformations may occur in an irregular way. In other words,
3D wire-printing has a morphological variability of its own.
Fig. 10. Diagram representing the sound proximity of each replica comparing to the
control flute, for each finger hole (numeric scale in semi-tones). The 0 line represents
the control flute. The colored areas represent the replicas’ sounds. The more the colored
area fills the 0 line, the closest the replica is to the control flute.
2. 3D resin-printing on the other hand does not work the same and thus does
not have the same sources of error [19]: it uses a laser impact which solidifies a
gelatinous resin. This technology is more accurate than 3D wire-printing and
gives different physical results (smoother state of surface, solid 3D printings).
That explains why this replica is much more accurate than the other ones.
3. Once again, human blow should be replaced by a mechanical blower.
6 Conclusion
Handcrafting and 3D replication processes illustrate the acoustical complexity of
bone flutes, as well as they raise most important epistemological and methodolog-
ical issues. Succinctly, these results advise of the dangers of sound-interpretations
regarding ancient flutes when dealing with replicas. They demonstrate the com-
plexity of the acoustical phenomenons related to naturally hollow raw-materials.
They also demonstrate that 3D imagery is not as precise and trustworthy as we
would think it would be. However, the use of statistics and of high-precision
3D printers seems to offer a promising track to continue this research. Although
there is still much work to do in order to reach a better understanding of this
situation, at least we now know that archaeological bone flutes sounds should
194 E. Safa et al.
always be interpreted with caution. In any case, this research will try and go
deeper in the epistemological and methodological issues.
Acknowledgments. This project was partially funded by the french CNRS ImagIn
IRMA project.
References
1. Avanzini, F., Canazza, S., De Poli, G., Fantozzi, C., Pretto, N., Roda, A.,
Menegazzi, A.: Archaeology and virtual acoustics - a pan flute from ancient Egypt.
In: Proceedings of the 12th International Conference on Sound and Music Com-
puting, pp. 31–36 (2015)
2. Barreau, J.B., Gaugne, R., Bernard, Y., Le Cloirec, G., Gouranton, V.: The West
Digital Conservatory of Archaelogical Heritage Project, DH, Marseille, France.
Digital Heritage International Congress, pp. 1–8, November 2013
3. Bellia, A.: The virtual reconstruction of an ancient musical instrument: the aulos
of selinus. In: Proceedings of Digital Heritage, vol. 1, pp. 55–58 (2015)
4. Borman, T., Stoel, B.: Review of the uses of computed tomography for analyzing
instruments of the violin family with a focus on the future. J. Violin Soc. Am. VSA
Papers 22(1), 1–12 (2009)
5. Brade, C.: Die mittelalterlichen Kernspaltflöten Mittel-und Nordeuropas, Ein
Beitrage zur Überlieferung prähistorischer und zur Typologie mittelalterlicher
Kernspaltflöten, Band. 14, Neumünster, Wachholtz (1975)
6. Clodor-Tissot, T.: Fiche témoins sonores du Néolithique et des Âges des métaux,
Industrie de l’os préhistorique. Instruments sonores du Néolithique à l’aube de
l’Antiquité, XII (2009)
7. Causs, R., Mille, B., Piechaud, R.: Restitution sonore numérique des cornua de
Pompei, Sound Making: Handcraft of Musical Instruments in Antiquity, video
recordings of the third IFAO conference, IRCAM (2016). http://medias.ircam.fr/
x27292e
8. Garca, B.C., Alcolea, M., Mazo, C.: Experimental study of the aerophone of Istu-
ritz: Manufacture, use-wear analysis and acoustic tests. Quaternary International
(2015). doi:10.1016/j.quaint.2015.11.033
9. Garca, B.C.: Methodology for the reconstruction of prehistoric aerophones made
of hard animal material, Actas das IV Jornadas de Jovens em Investigação
Arqueológica, pp. 411–416 (2011)
10. Gattoni, F., Melgara, C., Sicola, C., Uslenghi, C.M.: Unusual application of com-
puterized tomography: the study of musical instruments. Radiol. Med. 97(3), 170–
173 (1999)
11. Hagel, S.: The meroë pipes: a musical jigsaw puzzle, sound making: handcraft of
musical instruments in antiquity, video recordings of the third IFAO conference,
IRCAM (2016). http://medias.ircam.fr/x9e8e19
12. Hawkins, J., Jacobson, J., Franklin, J.: Greco-Roman Music in Context; Bringing
Sound and Music to Virtual Pompeii, World Conference on E-Learning in Corpo-
rate, Government, Health Care, and Higher Education (2011)
13. Koumartzis, N., Tzetzis, D., Kyratsis, P., Kotsakis, R.G.: A new music instrument
from ancient times: modern reconstruction of the greek lyre of hermes using 3D
laser scanning, advanced computer aided design and audio analysis. J. New Music
Res. 44(4), 324–346 (2015)
14. Kunej, D., Turk, I.: New perspectives on the beginnings of music: archaeological
and musicological analysis of a middle Paleolithic bone “flute”. In: The Origins of
Music, pp. 235–268 (2000)
15. Laycock, S., Bell, G., Mortimore, D., Greco, M., Corps, N., Finkle, I.: Combining
x-ray Micro-CT technology and 3D printing for the digital preservation and study
of a 19th century cantonese chess piece with intricate internal structure. ACM J.
Comput. Cultl. Heritage 5(4), 1–7 (2012)
16. Münzel, S.C., Seeberger, F., Hein, W.: The Geissenklösterle flute: discovery,
experiments, reconstruction, Studien zur Musikarchäologie. Archäologie früher
Klangerzeugung und Tonordnung, Rahden, Verlag M. Leidorf, pp. 107–118 (2002)
17. Pierce, L.B.: Hacking the Digital Print: Alternative image capture and printmaking
processes with a special section on 3D printing, New Riders, Voices That Matter,
p. 262 (2015)
18. Safa, E.: Handcrafting in archaeomusicological research: record of a one-year
apprenticeship alongside a traditional-flute-maker and its application to sound
archaeology, Sound Making: Handcraft of Musical Instruments in Antiquity, video
recordings of the third IFAO conference, IRCAM (2016). http://medias.ircam.fr/
xcf6cf9
19. Sculpteo: 3D Printing Material: PolyJet Resin, Sculpteos website (2016). https://
www.sculpteo.com/en/materials/polyjet-resin-material/
20. Sirr, S.A., Waddle, J.R.: CT analysis of bowed stringed instruments. Radiology
203(3), 801–805 (1997)
21. Sodini, N., Dreossi, D., Chen, R., Fioravanti, M., Giordano, A., Herrestal, P.,
Zanini, F.: Non-invasive microstructural analysis of bowed stringed instruments
with synchrotron radiation X-ray microtomography. J. Cultl. Heritage 13(3), S44–
S49 (2012)
22. Tuniz, C., Bernardini, F., Turk, I., Dimkaroski, L., Mancini, L., Dreossi, D.: Did
Neanderthals play music? X-ray computed micro-tomography of the Divje Dabe
’flute’. Archaeometry 54(3), 581–590 (2012)
23. Turk, I., Blackwell, B.A., Turk, J., Pflaum, M.: Results of computer tomography of
the oldest suspected flute from Divje bab I (Slovenia) and its chronological position
within global palaeoclimatic and palaeoenvironmental change during last glacial.
Anthropologie 110(3), 293–317 (2006)
24. Tzevelekos, P., Georgaki, A., Kouroupetroglou, G.: HERON: a zournas digital
virtual musical instrument. In: Proceedings of the 3rd International Conference
on Digital Interactive Media in Entertainment and Arts, pp. 352–359, September
2015
Full Paper: Digital Applications for
Materials’ Preservation and
Conservation in Cultural Heritage
Evolution of Building Materials
and Philosophy in Construction: A Process
of Digitalization and Visualization
of the Accumulated Knowledge
Ioanna Papayianni and Vasiliki Pachta(&)
Laboratory of Building Materials,

Department of Civil Engineering, Aristotle University Thessaloniki,
Thessaloniki, Greece
{papayian,vpachta}@civil.auth.gr
Abstract. The long-term research on the constructional materials and tech-

niques of monuments and historic buildings, allowed the accumulation of sig-
nificant knowledge which could be further disseminated. The masons of
antiquity followed principles in designing and building, established by their
intuition and experience. The selection of raw materials, the way they upgraded
them in constructing foundations, walls, domes, is still remarkable. In the paper,
a process of using digital technology tools for making knowledge acquisition
attractive is presented. By developing a specific platform, all relevant scientific
knowledge can be sorted, while with a series of digital applications, the dia-
chronic principles of construction, the ancient technology and the achievements
of the past can be exploited in a friendly and interactive environment. By this
way it is expected that the values of building philosophy in the context of safety,
sustainability and economy will be forwarded to new generations.
Keywords: Monumental heritage Building materials Constructions

Evolution Mortars
1 Introduction
It has been well established by the history of humanity that civilization proceeds in
parallel to building materials’ innovations. From the very first time man built a shelter
for protection, he realized that he could improve his life. He slowly, by practical
experience, acquired the necessary knowledge and elementary principles of building.
He used locally available materials (stone, clay) from nearby deposits and tried to treat
or mix them with straw and sand to make them more resistant to natural environment.
For centuries, masons invested in details and created some of the master pieces of
architecture and engineering, such as Pantheon and Hagia Sophia, still alive.
The selection of the topography of the place and the geometry of the buildings’
ground plan were carefully made, so as the structure to cooperate with natural envi-
ronment by arranging the openings for ventilation and lightening.

DOI: 10.1007/978-3-319-48496-9_16
200 I. Papayianni and V. Pachta
These are some of the principles followed in the past constructions. Studying the
materials of monuments, a treasure of knowledge and masons’ dexterities is revealed.
For example, in ancient Olynthos (4th century BC) they used mud mortar for the
construction of stone walls, but they plastered the walls for functioning in moisture
with thick lime-pozzolan mortar plasters (Fig. 1) [1].
Fig. 1. Macroscopic and microscopic figure of plaster from the archaeological site of Olynthos
Some writers of antiquity like Vitruvius and Plenius [2] have included in their
treatises precious practical instructions on preparing building materials and using them,
but after Roman period the knowledge transfer was limited [3].
During the recent past, for the needs of restoration, the ancient materials were
studied and analyzed systematically and many philosophical principals on which
building was based were recognized [4–6]. At the same time, huge data bases have
been developed at Research Centers containing real treasures of knowledge concerning
building materials, without any beneficial for the society exploitation [6–8].
Nowadays, the advances in building materials (high performance and strength
concrete) lead to the construction of skyscrapers, expressing the spirit of modern
society. However, under the prism of sustainability (saving energy, reducing envi-
ronmental footprint, recycling), building materials and technology are reconsidered.
The science of building materials provides the knowledge at all levels of controlling
and increasing the potential of them and allows understanding how materials work in
the structure.
While architecture in construction has drawn the interest of society, building
materials and technological evolution have not appeared in museum exhibitions at
European level and all concepts and philosophy of building have not been highlighted
to gain educational value, to offer multi-level and multidisciplinary readings and inspire
new generations for a better future in building.
On the other hand, there is an increase of research interest on the application of
Information and Communications Technology (ICT) tools in Cultural Heritage, such as
Historic Building Information Modelling (HBIM), development of virtual museums
and exhibitions, visualization of objects and artifacts, development of 3D environments
and experiences [9–11].
Evolution of Building Materials and Philosophy in Construction 201
Under these aspects, the paper focuses on an attempt to gather, evaluate and finally
visualize scientific results on the building materials and techniques used in construc-
tions during the last four millenniums, starting from prehistory until nowadays. The
study aims at covering a significant gap of knowledge, regarding the technological
evolution of structures and developing a dynamic platform by processing, dissemi-
nating and promoting knowledge regarding:
• The significant role of building materials in the development of the architecture and
civilization of an area
• The evolution of the building materials technology from the past (2nd millennium
BC) until nowadays (advances of concrete)
• The diachronically valid principles of sustainability, durability and economy in
construction
• The proper materials and techniques for the restoration of monuments and historic
buildings.
2 Principles and Objectives
ICT (information and communications technology) is an umbrella term including many

applications and during the last decade is interrelated with multiple research sectors in
order to promote and disseminate scientific knowledge [12–14].
The use of ICT tools in understanding and promoting specific aspects of Cultural
Heritage has been studied by many researchers [15–17] and applied by various
Organizations, museums and galleries. On the other hand, the need of digitalizing and
visualizing archives and collections is increased.
The idea of developing of a dynamic platform on building materials and techniques
from the past up to nowadays emerged from the weakness of relevant museums of
Architecture or monuments’ exhibition to cover material’s aspects, the inadequacy of
education on these topics, as well on the lack of finding concentrated scientific
knowledge on the evolution of constructional technology.
The invention of cement and the continuous evolution of concrete technology of
nowadays, demonstrate the close relation of modern society with building materials and
technology and lead to the adoption of the basic environmental issues (i.e. saving
energy, reducing environmental footprint, recycling in construction).
Furthermore, it is of great importance to make people think and be conscious about
building, since construction is the sector that mostly influences the economy of natural
sources and environment, in other words the future of humanity. If people choose for
building their houses local materials of high ecological profile, the CO2 footprint of
human activities could be reduced.
The proposed platform aims at managing, exploiting and disseminating a large
volume of scientific knowledge regarding the evolution, characteristics and effective-
ness of different building materials and technologies. Earth, stone, bricks, limes,
pozzolans, cements, mortars and concretes that have dominated in housing for four
millenniums are addressed, showing the entity of building the space with materials
taken from the Earth.
The innovation of the platform is that such a museum dedicated to the common
European building materials and technology, does not exist either in virtual or in
conventional form. In relevant contemporary museums and exhibitions of architecture,
there are very limited references to the building materials and technology of specific
historic periods and regions. Therefore, a holistic approach on the evolution of building
materials and technology is not provided to visitors. Philosophical aspects (such as
functionality, sustainability) and their diachronicity in construction are also not
revealed to visitors.
3 Process Followed
The proposed dynamic platform will provide the opportunity for visitors to acquire,
extend or deepen their knowledge on building materials and techniques, through
story-telling authoring tools and digital creations such as virtual reality and 3D
experiences.
Different aspects on the technology of building materials, such as structural types,
selection of raw materials, manufacture and application of materials, properties and
characteristics, damages and repair of structures, will be presented under the prism of
sustainability and economy in construction, covering a historical period of more than 4
millenniums.
Scientific knowledge on building materials will be combined with ICT technology,
in order to facilitate the access of all levels of users. Best practices will be presented,
highlighting that the science of building materials provides the knowledge of con-
trolling and increasing their potential, as well as understanding how materials work in
each structure.
It is expected that a consciousness about cost-effective and “green” building will be
developed in the users of the coming generations.
The Process followed could be distinguished in specific tasks, including:
• Management of a large volume of scientific knowledge, regarding research analysis
results of building materials (mortars, stones, bricks, concrete technology) and
critical review of relevant literature concerning the types and characteristics of
building materials used in constructions for millenniums.
• Development of a flexible and easily comprehensive platform for the management
and dissemination of all information that could be moderated according to future
needs. The platform will have a dual performance, referring to the input of all data
and the narrative and virtual representation of the information to users, according to
their needs.
• Visualization of all data through the development of architectural designs, structural
representations, 3D animations and videos. Story-telling digital tools will be used
for the dissemination of knowledge. For example, interactive maps will indicate
structures of specific type (i.e. adobe masonry) with emerging information on
historic, architectural and morphological issues (Fig. 3). Narrative three dimen-
sional environments with specialized information (in different forms, such as dia-
grams, tables, photos) regarding the characteristics and properties of specific
building materials and structural systems will be developed.
The schematic diagram of the process is depicted in Fig. 2.
Fig. 2. Schematic diagram of the process followed
Fig. 3. Interactive map with indicative structural types. Sanctuary of Dionysus, archaeological
site of Maronia (4th cent. BC)
4 Data Input
As it has been described, the evolution of building materials and techniques has not
been adequately interrelated with the growth of civilizations and architecture, although
all innovations and achievements in building technology have been implemented with
materials. Therefore, there is a significant lack of knowledge regarding the role of
structural materials in developing European society and economy.
Moreover, the strong influence of built environment to the nature has not been
sufficiently understood by society. The comprehension of the philosophy of building in
the past highlights the principles of building, such as how the local environment was
taken into account, as well as the economy in using natural resources.
Under the prism of sustainability, building materials and technology are nowadays
reconsidered, since it has been well proved by the experience of great catastrophes due
to climatic change that the sector of constructions is the major consumer of natural
resources and sustainable building is one way road to limit these phenomena. As a
result, the research on upgrading local building materials is cutting edge technology
and billions of euros are annually invested to that direction.
For example, earth structures which concern the primer constructional system
found, have been continuously used in construction from prehistory until nowadays
(33 % of worldwide houses are still built with earth), due to their low cost and easy
production, without high energy embodied materials. They were constructed with local
clayish materials by using techniques which depicted the regional constructional tra-
ditions (earth-blocks in SE Europe, cob in UK and taipa in Portugal) (Fig. 4).
Fig. 4. Earth-block houses of N Greece
Under the light of sustainability, these structures have been reconsidered and efforts
are made in order to be restored and revitalized with proper materials and techniques.
On the other hand, there is a revival of the interest on building with earth all over the
world (i.e. France, Germany), due to the constructional, economic and environmental
benefits [20].
Fig. 5. Various structural systems used in constructions
Through the platform these aspects will be highlighted in order to offer multidis-
ciplinary readings and inspire users for a better future in building. Apart from
increasing the dissemination potential of knowledge about the evolution of building
materials, the scientific experience and knowledge will be presented under the concepts
of effectiveness, economy, sustainability and environmental protection.
Well documented knowledge has been accumulated by the Laboratory of Building
Materials of the Aristotle University of Thessaloniki during the last 25 years of con-
tinuous work with Ephorates of Antiquities in studying the building materials of his-
toric structures [4, 6, 18, 19].
According to a systematic evaluation of the analysis results, many conclusions can
be asserted regarding multiple aspects, such as the type of building materials used
diachronically, the constructional types encountered, as well as their correlation to the
specific characteristics of each era and region (socio-economic, cultural and environ-
mental aspects).
In the proposed platform, the research results will be presented under specific axes,
regarding:
• the main structural types found diachronically in constructions (rubble masonry,
ashlar masonry, adobe, brick masonry, contemporary concrete structures) (Fig. 5)
• the characteristics and technological evolution of building materials and techniques
(Fig. 6)
• the wider historic, cultural, environmental and socio-economic aspects of each
region and era
• the background and educational level of the user (general public, students, pro-
fessionals, scientists, stake holders).
Fig. 6. Macroscopic and microscopic figures of mortars from various monuments of Greece
5 Conclusions
As it is formerly stated, the technological evolution of building materials has not been
yet adequately acknowledged and interrelated with the development of civilizations
and advances in construction. Therefore, there is a gap of knowledge related to the
significant role of building materials and philosophy of construction in creating the
built environment in Europe.
On the other hand, under the prism of sustainability, the comprehension of the
diachronic principles of constructions could contribute to maximizing the effectiveness
of the contemporary building materials and minimizing their cost and environmental
footprint.
To this direction, the development of a flexible platform providing input to many
digital applications concerning building materials and constructional techniques would
contribute to forward the messages to the next generations:
• the significant role of building materials in the development of the architecture and
civilization of a region,
• the evolution of the building materials technology from the past (2nd millennium
BC) until nowadays (advances of concrete),
• the diachronically valid principles of sustainability, durability and economy in
construction,
• the proper materials and techniques for the restoration of monuments and historic
buildings.
References
1. Papayianni, I., Stefanidou, M.: Durability aspects of Ancient Mortars of the archeological
site of Olynthos. J. Cult. Heritage 8, 193–196 (2007)
2. Morgan, M.H.: Vitruvius: The Ten Books of Architecture. Dover Publications Inc.,
New York (1960)
3. Palladio, A.: The Four Books of Architecture. Dover Publications Inc., New York (1965).
Unabridged republication of the work first published by Ware I in 1938
4. Papayianni, I.: The longevity of old mortars. Appl. Phys. A Mater. Sci. Process. 83, 685–688
(2006)
5. Middendorf, B., Hughes, J.J., Callebaut, K., Baronio, G., Papayianni, I.: Investigative
methods for the characterisation of historic mortars-Part 1: mineralogical characterization.
Mater. Struct. 38, 761–769 (2005). RILEM TC 167-COM: ‘Characterisation of Old Mortars
with Respect to their Repair’
6. Pachta, V.: Study of the Technological Evolution of Mortars, Dissertation. Aristotle
University of Thessaloniki, Thessaloniki (2011)
7. Papayianni, I., Pachta, V., Iliadou, K.: A data base system for managing information
concerning historical mortars. In: Ioannides, M., Addison, A., Georgopoulos, A.,
Kalisperis L. (eds.) VSMM 2008, pp. 271–277 (2008)
8. Moropoulou, A., Polikreti, K., Bakolas, A., Michailidis, P.: Correlation of physicochemical
and mechanical properties of historical mortars and classification by multivariate statistics.
Cem. Concr. Res. 33, 891–898 (2003)
9. Athanasiou, G.S., Michail, H.E., Gregoriades, A., Ioannides, M.: Evolution of the e-Museum
concept through exploitation of cryptographic algorithms. In: Ioannides, M., Fritsch, D.,
Leissner, J., Davies, R., Remondino, F., Caffo, R. (eds.) EuroMed 2012. LNCS, vol. 7616,
10. Kiourt, C., Koutsoudis, A., Pavlidis, G.: DynaMus: a fully dynamic 3D virtual Museum
framework. J. Cult. Heritage. http://dx.doi.org/10.1016/j.culher.2016.06.007
11. Robles-Ortega, M.D., Feito, F.R., Jiménez, J.J., Segura, R.J.: Web technologies applied to
virtual heritage: an example of an Iberian Art Museum. J. Cult. Heritage 13, 326–331 (2012)
12. Ashlock, D.: Evolutionary Computation for Modelling and Optimization. Springer,
New York (2006)
13. Comby-Peyrot, I., Bernard, F., Bouchard, P.O., Bay, F., Garcia-Diaz, E.: Development and
validation of a 3D computational tool to describe concrete behaviour at mesoscale.
application to the alkali-silica reaction. Comput. Mater. Sci. 46, 1163–1177 (2009)
14. Donga, S., Wang, X., Xu, S., Wu, G., Yin, H.: The development and evaluation of Chinese
digital science and technology Museum. J. Cult. Heritage 12, 111–115 (2011)
15. Donga, S., Xu, S., Wu, G.: Earth Science Digital Museum (ESDM): toward a new paradigm
for Museums. Comput. Geosci. 32, 793–802 (2006)
16. Pavlidis, G., Koutsoudis, A., Arnaoutoglou, F., Tsioukas, V., Chamzas, C.: Methods for 3D
digitization of cultural heritage. J. Cult. Heritage 8, 93–98 (2007)
17. Sylaiou, S., Liarokapis, F., Kotsakis, K., Patias, P.: Virtual Museums, a survey and some
issues for consideration. J. Cult. Heritage 10, 520–528 (2009)
18. Pachta, V., Stefanidou, M., Konopissi, S., Papayianni, I.: Technological evolution of historic
structural mortars. J. Civ. Eng. Architect. 8(7), 846–854 (2014)
19. Stefanidou, M., Papayianni, I., Pachta, V.: Evaluation of inclusions in mortars of different
historic periods from Greek monuments. Archaeometry 54, 737–751 (2012)
20. Papayianni, I., Pachta, V.: Earth block houses of historic centers. a sustainable upgrading
with compatible repair materials. Procedia Environ. Sci. (in press)
A Study of 3D Digital Simulation Analysis
of Fire Charring Degree of Wood Construction
of Chinese Traditional Architecture
Tsung Chiang Wu(&)
Department of Civil Engineering and Engineering Management,

National Quemoy University, 89250 Kinmen, Taiwan
[email protected]
Abstract. For the Chinese traditional architecture which uses wood construc-
tion in large quantities, the fire often causes irreversible disasters, and the cul-
tural heritage may be lost in a flash. According to Taiwan Cultural Assets
Preservation Act, the historic monuments restoration must uphold the spirit of
“Restoring the Old as the Old”, so the structural safety assessment is a necessary
program for what can be restored after disaster. Traditionally, the char depth
data of the wood construction after fires are obtained by pore-drilling mea-
surement. Therefore, the detection positions and quantity are determined
according to the post-disaster condition. The weak structure often fails to be
measured, so that the evaluation result is likely to be distorted. This study uses
dynamic fire simulation theory, and takes a Chinese traditional architecture,
Potzu Pei-tian Temple in Chiayi, Taiwan as an example for experiment, trying to
build a digital char depth virtual detection model. The research findings show
that this conception can build a virtual detection mode, which may provide more
comprehensive char depth information than traditional method, assisting the
safety assessment operation of post-disaster restored structures effectively.
Keywords: Digital simulation Charring degree Dynamic fire simulation

theory Virtual detection mode
1 Introduction
The “fires” will cause irrecoverable results of cultural heritage preservation and
inheritance, the intelligence and sequence of thought transferred by ancient people for
hundreds of years may be destroyed and eliminated in a short period of time. However,
if a fire happens unfortunately, for the restorable case, the situation shall be assessed
efficiently and accurately before related restoration plans are made, so as to avoid
further losses. Before dawn on March 26, 2013, the 330 years old historic monument,
Potzu Pei-tian Temple of Chiayi, Taiwan fired, the wood construction-based main hall
was burnt down, including numerous historic relics, the loss was heavy. It is so
regrettable that on November 15, 2012, a grand ceremony for the completion of
Pei-tian Temple which cost 4 years and NT$120 million was held. The fire burnt down
the four years’ efforts, and caused irrecoverable losses of Taiwan’s cultural heritage. In
order to push the subsequent maintenance work, the post-disaster on-scene safety

DOI: 10.1007/978-3-319-48496-9_17
210 T.C. Wu
assessment is the first work. Therefore, the structural safety of the burnt heavy timber
construction and the charring degree of beam columns must be further detected. The
old charring degree detection methods used tube sample boring for actual measure-
ment, related samples were insufficient, or the positions were inappropriate, the eval-
uation result was likely to be distorted. If there are too many holes drilled, there is a
high risk of secondary damage to the post-disaster remnant. Therefore, how to take
sufficient samples and reduce the risk of secondary damage shall be thought over and
studied.
This study tries to integrate digital building and digital fire simulation techniques
for virtual simulation of the process of fire. The wood construction is taken as an
example, referring to the experience in the study of charring rate estimation model [1]
and the study of wood characteristics [2], the temperature change withstood by the
wood construction in the fire simulation results is analyzed. Afterwards, the charring
degree in relative position of wood construction is estimated according to the “rela-
tionship between temperature and charring degree” established in laboratory [3]. In this
conception, the spatial pattern of the distressed architectural space is built by 3D digital
modeling, and the fire is simulated by FDS (Fire Dynamics Simulator) software. The
3D realistic fire simulation technique will be used to preliminarily detect and estimate
the wood construction char depth of the distressed building, and the quantity of drill
samples may be reduced, so as to analyze the wood construction char depth rapidly.
2 Method
Potzu Pei-tian Temple, Chiayi, Taiwan is taken as an example for virtual simulation
and detection of the char depth of the wood construction of the burnt main hall space.
The research process and the outcomes are detailed below:
2.1 Collect and Integrate the Data of Study Case

This study case is Potzu Pei-tian Temple, Chiayi, Taiwan (Fig. 1 left). The temple is
registered as a county historic monument of Taiwan, it has very high historic signifi-
cance and cultural value. Pei-tian Temple was founded in the 26th year of Kangxi of
the Qing Dynasty (1687). It was built for the fourth time during the Japanese occu-
pation in 1915, under the charge of Ying-bin Chen, a famous master from Zhangzhou,
China. The “opposite construction” was adopted, and it was the first temple in Taiwan
subsidized by Japanese authority [4]. The latest construction was completed in 2012,
but the main hall was burnt down in 2013 (Fig. 1 right). In order to probe into the
historical evolution and building maintenance history of the study case, the historical
documents, investigation reports, planning and design, rehab works reports, architec-
tural drawings and records of interview with the parties concerned about the temple are
collected.
A Study of 3D Digital Simulation Analysis of Fire Charring Degree 211
Fig. 1. Potzu Pei-tian Temple, Chiayi, Taiwan (left), main hall burnt (right)
2.2 Build Spatial 3D Digital Model of the Burnt Main Hall

As the scene of the burnt main hall was messy, there was no spatial information
obtained from the scene. Therefore, the 3D digital model of the burnt main hall was
built (Fig. 2 right) by using SketchUp according to the collected architectural drawing
(Fig. 2 left) and related image data as the FDS software.
Fig. 2. Architectural drawing (left), 3D digital model of main hall (right)
2.3 Construction of Attribute Data of Burnt Main Hall

The materials and size of the space must be given for FDS. Therefore, the material and
size of the main hall wood construction are given referring to related architectural
drawing, investigation and research reports and restoration design planning data. The
wood construction material of this study case is fir.
2.4 FDS Test

For accurate FDS, the input conditions must match current condition as possible, and
the simulation environment shall be tested by appropriate 3D simulation grid and the
212 T.C. Wu
digital thermometer shall be selected and mounted [5] before FDS test. The settings of
environmental parameters must match the research purpose. As the FDS space in this
study is an indoor confined space, when the FDS is used for calculation, relevant
parameters must be set. This step analyzes the effect of the precision of spatial layout
and size of fire scene on the fire simulation result, so the spacing between digital space
nodes is set as 25 cm in x, y and z directions. According to the equation Q ¼ at2 of
“t-squared fires” [6], where Q is the heat release rate, a is the fire burning rate, the
growth time t is defined as the time for active combustion to grow to 1; 055 kw
ð1; 000 Btu/sÞ. The heat release rate of the fire source at initial stage of fire is set
according to t2 swift growth, the burning rate a is 0:01689 kw/s2 , the peak of fire heat
release rate is 10; 000 kw [7, 8]. After the aforesaid data are imported and the envi-
ronmental parameters are set, the FDS is implemented. The total time of the simulated
burning process is accelerated, the simulation is finished in 10 min. The simulated
burning process is shown in Fig. 3 including 420 s, 480 s, 540 s and 600 s of
combustion.
Fig. 3. FDS simulating process
2.5 Analysis of Wood Construction Component Temperature Change

FDS provides the function of digital thermometer, which can be mounted on the 3D
digital model surface. There are 14 digital thermometers located on the main hall
structure components, coded A1, A2, A3, B1, B2, B3, C1, C2, C3, D1, D2, D3, E1 and
G1. The locations are shown in Fig. 4. The temperature-time change values of digital
thermometers on the wood construction are obtained after FDS. The temperature
Fig. 4. 14 digital thermometers located on main hall components
changes of wood construction, such as beams and columns of main hall, in the fire
simulation process are analyzed. The maximum combustion temperatures in the
positions detected by the 14 digital thermometers are listed in Table 1.
Table 1. Maximum temperatures recorded by 14 digital thermometers

Position Detected maximum Position Detected maximum
code temperature (°C) code temperature (°C)
A1 460 C1 1110
A2 460 C2 1120
A3 460 C3 465
B1 1180 D1 465
B2 1225 D2 470
B3 460 D3 470
E1 470 G1 1240
2.6 Comparative Analysis of Charring Data

Referring to the look-up table of CNS 12514 heating temperature and time (Table 2),
the regression curve of heating temperature and time is calculated (Fig. 5), the tem-
perature and time are calculated, the burning duration in 14 positions is obtained. The
regression equation is calculated according to the look-up table of CNS 12514 heating
char depth and charring rate of fir material in Table 3 (Fig. 6), the burning duration,
char depth and charring rate are compared, so as to analyze the values of char depth and
charring rate.
214 T.C. Wu
Table 2. Look-up table of CNS 12514 heating temperature(TP, °C) and time(TM, sec.)
TM TP TM TP TM TP TM TP TM TP
1 349 13 717 25 815 37 873 49 915
2 445 14 728 26 820 38 877 50 918
3 502 15 739 27 826 39 881 51 921
4 544 16 748 28 832 40 885 52 924
5 576 17 757 29 837 41 888 53 927
6 603 18 766 30 842 42 892 54 930
7 626 19 774 31 847 43 896 55 932
8 645 20 781 32 851 44 899 56 935
9 663 21 789 33 856 45 902 57 938
10 678 22 796 34 860 46 906 58 940
11 693 23 802 35 865 47 909 59 943
12 705 24 809 36 869 48 912 60 945
Fig. 5. CNS 12514 standard heating temperature and time regression curve and equation
Table 3. Look-up table of fir heating time and char depth of CNS 12514 [9]
Tree species Heating time (sec.) Char depth (mm)
Fir 1800 21.3(2.6)
2700 26.8(3.6)
3600 43.4(5.7)
Brackets(): standard deviation
Fig. 6. Regression curve and equation of fir heating time and char depth
3 Discussion
The outcomes obtained by the wood construction char depth simulation detection
model built in this study are shown in Table 4. The detection positions B1, B2 and G1
are nearby the fire source, receiving higher temperature, so the char depths are greater
than 41.55 mm; the char depths in the other positions are 10.00 mm * 11.00 mm.
Table 4. Detection results of wood construction char depth simulation detection model
Detection Simulated combustion Corresponding burning Corresponding char
position temperature (°C) duration (sec) depth (mm)
A1 460 135.6 10.57
A2 460 135.6 10.57
A3 460 135.6 10.57
B1 1180 >3600.00 >41.55
B2 1225 >3600.00 >41.55
B3 460 135.6 10.57
C1 1110 >3600.00 >41.55
C2 1120 >3600.00 >41.55
C3 465 143.4 10.61
D1 465 143.4 10.61
D2 470 151.2 10.61
D3 470 151.2 10.61
E1 470 151.2 10.61
G1 1240 >3600.00 >41.55
216 T.C. Wu
4 Conclusion
The detection of char depth of wood construction after fires is a necessary procedure of
structural safety assessment. The non-contact digital simulation can avoid the drilling
causing secondary damage to the historic monuments and historical buildings, and the
risk of assessors in work can be reduced. The experimental results show that the digital
simulation detection model built in this study can assist in the structural safety
assessment of historic monuments and historical buildings after fires, and can make up
the deficiencies in char depth evaluation by drill hole sampling, so as to perfect the
evaluation operation. As the wood construction only uses wood as sample parameter, if
the actual wood species parameter of analysis can be provided and the water content
and surface material (e.g. paint, colored drawing) are considered, the model built in this
study will be sophisticated.
Acknowledgments. This research was supported by the Taiwan Ministry of Science and
Technology “A study of analyzing carbonized degree of historical sites’ timbers using 3D
realistic simulation of fire scene” (Project No. 103-2410-H-507-006-).
References
1. White, R.H., Dietenberger, M.A.: Wood handbook-wood as an engineering material,
Chap. 17. In: Fire Safety. Forest Products Laboratory (1999)
2. Njankouo, J.M., Dotreppe, J.-C., Franssen, J.-M.: Experiment study of the charring rate of
tropical hardwoods. Fire Mater. 28(1), 15–24 (2004)
3. Schaffer, E.: State of structural timber fire endurance. Wood Fiber Sci. 19(2), 145–170 (1977)
4. Bureau of Cultural Heritage, MOC, Taiwan website. http://www.boch.gov.tw/
5. McGrattan, K., Klein, B., Hostikka, S., Floyd, J.: Fire: dynamics simulator (Version 5) user’s
guid. NIST Spec. Publ. 1019–5 (2007)
6. Milke, J.A.: Smoke management for covered malls and atria. Fire Technol. 26(3), 223–243
(1990)
7. Spearpoint, M.J.: Predicting the ignition and burning rate of wood in the Cone Calorimeter
using an integral model. NIST GCR, 99–775 (1999)
8. Spearpoint, M.J., Quintiere, J.G.: 2001, Predicting the ignition of wood in the Cone
Calorimeter - effect of species, grain orientation and heat flux. Fire Saf. J. 36(4), 391–415
(2001)
9. Wang, S.-Y.: Fire behavior and fire protection procedures for wooden historical buildings.
J. Cult. Property Conserv. 1(1), 18–26 (2007). (in Chinese)
Full Paper: Non-destructive Techniques
in Cultural Heritage Conservation
Recovering Historical Film Footage
by Processing Microtomographic Images
Chang Liu1,2 , Paul L. Rosin2(B) , Yu-Kun Lai2 , Graham R. Davis3 ,

David Mills3 , and Charles Norton4
1
School of Astronautics, Beihang University, Beijing, China
2
School of Computer Science & Informatics, Cardiff University, Cardiff, UK
[email protected]
3
Institute of Dentistry, Queen Mary University of London, London, UK
4
BBC Archive Development, London, UK
Abstract. 1960s film was typically printed on tri-acetate film base. If

not preserved properly, such material breaks down at a chemical level,
which is a non-stoppable process that permanently fuses the film so that
it essentially becomes a lump of solid plastic. Recently, some precious
films, such as the only known copy of the earliest surviving episode of
‘The Morecambe and Wise Show’ have been discovered, but they are
in poor condition. They will eventually turn into a pool of sticky liquid
and be gone forever. In this paper, as proof of concept, we use X-ray
microtomography to provide 3D imaging of a test film of similar vintage,
and propose an automatic method to extract footage from it.
Keywords: Tri-acetate film · X-ray microtomography · 3D imaging
1 Introduction
In 1968, Eric Morecambe and Ernie Wise came to London to make a new light
entertainment series for BBC 2, which would come to dominate the UK light
entertainment landscape for the next 15 years. Sadly, the BBC never formally
archived that very first season of ‘The Morecambe and Wise Show’. Before the
advent of commercial video exploitation (DVDs, iplayer etc.), bodies like the
BBC saw little purpose in maintaining a large and expensive collection of pro-
gramming. ‘The Morecambe and Wise Show’ (like almost all UK comedy broad-
casts of the time) was recorded to and transmitted from videotape. After being
broadcasted, the master-tape would be stored for a limited period of time for
potential re-use and eventually cleared for wiping.
A programme might still survive, but only if it was deemed likely to be
popular overseas. Around the time of broadcast, many BBC programmes would
also be copied to 16 mm black and white film via a process called telerecording. A
master-tape would be played on a special flat cathode ray tube screen. A 16 mm
film camera was pointed at this screen and filmed the programme. These film
recordings were made for international syndication, and, since most countries

DOI: 10.1007/978-3-319-48496-9 18
220 C. Liu et al.
were able to play 16 mm film, this became the standard format for the marketing
of BBC programmes overseas.
All of the episodes of that first season of ‘The Morecambe and Wise Show’
were telerecorded to black and white film. Rather than having multiple film
prints created for each episode, it was routine for prints to be shared. When the
episodes had been screened by everyone who wanted them, the last network in
the chain would be asked to send the film back to BBC Enterprises in London or
for the films to be destroyed (to stop them being screened without permission).
By the late 1970s, series one of ‘The Morecambe and Wise Show’ had reached
Nigeria, in the film stores of a broadcaster called RKTV.
At some point prior to 1978 BBC Enterprises decided that there was no need
to retain their black and white negatives of ‘The Morecambe and Wise Show’
since colour television had become established across the world. They destroyed
the films, and along with the destruction of their master-tapes some years before,
this now meant that the BBC no longer held any recordings of the first series of
‘The Morecambe and Wise Show’.
Then in 2012, a freelance film researcher called Philip Morris gained rare
access to the RKTV vault, which consisted of a hot, poorly ventilated outbuilding
in a small settlement called Jos in the heart of Nigeria. The area suffers from
high-humidity in the rainy season and a desert of drought at all other times.
The often poor state of the films in the archive reflects these harsh conditions.
Philip Morris’ trip to the RKTV vault was worthwhile, as it turned out RKTV
held unique film recordings of various BBC programmes that were not known
to exist elsewhere in the world, including episodes of both ‘Doctor Who’, ‘The
Sky at Night’ and finally also ‘The Morecambe and Wise Show’. RKTV now
retained only one episode of ‘The Morecambe and Wise Show’, but it was still an
important find. On a single 16 mm film spool, RKTV had the earliest surviving
episode of ‘The Morecambe and Wise Show’ and the earliest surviving footage
of the double act working for the BBC.
However, there was a problem: the Nigerian Morecambe and Wise film was
in an advanced state of decay. The episode was on a form of film stock (i.e. the
flexible ribbon of transparent plastic backing onto which the image is printed) not
used nowadays called cellulose acetate, and was suffering from ‘tri-acetate film
base degradation’ – a film ‘disease’ that can cause film to rot away to nothing.
If kept in humid conditions for any prolonged period of time, the tri-acetate can
begin to break down at a chemical level, leading to a runaway reaction. At a
molecular level, as the plastic increases in temperature, the acetate breaks down
releasing a vapour of acetic acid, which further breaks down the plastic by eating
away at the film, through general acidic corrosion. The process only stops when
there is no more acetate left to be converted [4].
The RKTV Morecambe and Wise film had been stored in exactly the wrong
sort of conditions for an acetate roll: a humid atmosphere with poor ventilation
and a cracked film can that did not shield the film from sunlight. As the acid
has eaten away at the surface of the film, the roll has become sticky, thinner
and lacking in overall structural integrity. Every layer of film on the Morecambe
Recovering Historical Film Footage by Processing Microtomographic Images 221
and Wise roll has now permanently fused itself with every adjacent layer of film,
so that all of the pictures are now sealed inside a solid lump of plastic. Since
the film is sitting in a self-sustaining vapour of corrosive acid, eventually this
priceless roll will turn into soup. There is nothing we can do to stop that rot.
The BFI film archives recommended its destruction since the sheer quantity
of acid vapour being produced by the film made it a hazard to other rolls in the
BFI’s collection. The risk of ‘infection’ is always high with this kind of stock.
There is a long history of film restoration [1], and many techniques have
been developed [6,7]. Much of that work involves applying image processing
techniques to enhance existing footage, whereas for the Morecambe and Wise
film that footage needs to be first recovered from the acetate roll. In this paper
we describe a radically different approach to dealing with the film, inspired by
work in other areas of culture heritage.
In 2012, archaeologists embarked on a project to examine the Antikythera
mechanism – a Greek navigational calculator constructed around the end of the
second century BC. High-resolution X-ray tomography enabled them to analyse
the inner workings of the device, without breaking the item apart [5]. Recently,
several of the authors of the current paper have worked on the task of reading
historical scrolled parchment documents which have been damaged due to fire or
water. Such damage makes the document fragile, and often impossible to open
without causing considerable further damage. It was demonstrated that using
high-resolution X-ray tomography it was possible to virtually unroll a parchment
from the X-ray volume, and recover its written contents [9].
Any roll of film is essentially just a strip of plastic with a picture printed
onto one side of it, which for a monochrome image is always applied in the same
basic way. The image layer is called the emulsion layer and the key component
of black and white emulsion is silver oxide. The metal content in silver oxide
emulsion is very high. This suggested that it should be possible to apply X-ray
microtomography to obtain images from a roll of black and white film.
As a study of feasibility, in this paper we use the film of ‘Snow White and the
Seven Dwarfs’, a black and white film on tri-acetate stock of roughly the same
vintage as ‘Morecambe and Wise’. It was an 8 mm silent black and white digest
print from the 1970s containing the 1965 American dubbed version of the 1955
German film ‘Schneewittchen’. We first introduce how the film was scanned and
some early effort to construct the footage, which involved user interaction. We
then introduce a novel automatic algorithm to reconstruct the footage. This is
followed by experimental results and discussions.
2 Film Scanning and Early Effort of Reconstruction

Prior to tomographic scanning, a single thickness of 8 mm silent black and white
film (extracted from Snow White) was X-rayed at 40 kV and 405 µA, using a
0.5 mm Al filter. Although this X-ray energy was far too high for optimal con-
trast, it was necessary in order to penetrate even a small roll of film. The MuCAT
2 X-ray microtomography scanner at Queen Mary University of London [2] was
222 C. Liu et al.
used, which employs time delay integration to obtain very high contrast images.
But even with an equivalent exposure time of 40 s, little image contrast could
be seen. Given the poor image contrast, a long exposure time was set for the
tomographic scan. This was also performed at 40 kV and 405 µA. A length of
the film was rolled and inserted into a 27 mm internal diameter plastic container.
2115 projections were recorded with an equivalent exposure time of 20 seconds
per projection to ensure a high SNR. The voxel size was 20 µm.
The reconstructed film scan showed good signal to noise ratio. Streak arte-
facts were present, due to tangential X-ray paths passing through relatively long
lengths of silver. For future scans, these can be reduced in severity by increasing
the number of projections (the exposure time can be reduced to keep the same
total acquisition time).
The recent work in [3] demonstrated the possibility of recovering footage
from XMT scanned data. The method however requires manual interaction and
assumes uniform structure of the film. Moreover, it would be difficult to gener-
alise such methods to a highly damaged dataset.
3 Method for Image Recovery from Film

In this work, we will describe an automatic approach to footage recovery. We
first introduce the method to recover the image printed on the film and in the
next section, the method to create the video footage from such images.
3.1 Preprocessing and Initialisation

An example slice of the Snow White film is shown in Fig. 1. It looks similar to
parchments which our previous research has developed several effective methods
to deal with [8–10]. However, the methods for parchment unrolling fail on this
data because they assume that the parchment layers are mostly separated by at
least a small gap, whereas the film is more tightly rolled. On the other hand,
film is more uniform than parchment. Thus, in this work we focus on modelling
the appearance of the film, for which we use image intensity ridges [11], while
making no assumptions about any background gaps. Ideally, the ridges should
form a continuous spiral curve, but they are often incomplete. There also exist
artifacts, false positive responses which should be discarded. Based on the fact
that adjacent slices have similar images and hence similar ridge locations, we
semi-automatically correct the first slice, and automatically propagate the ridges
across slices, using the previous slice to constrain the linking of the next slice.
3.2 Propagation
Because of the large similarity between two adjacent images, if the ridge sections
in the previous image have been correctly linked, so that there exists an entire
ridge in the previous image, we can use the previous ridge image as a reference
to connect the ridge sections in the current image. In Fig. 2(a), the blue spiral
Fig. 1. An X-ray image of Snow White film. A section is shown zoomed in.
represents the previous ridge, i.e. the reference ridge; the red curves are some
ridge fragments in the current image. Generally, if a fragment is a section of the
film ridge rather than an artifact, it should be very close to the reference ridge.
By means of this condition, we check whether a fragment is a ridge part or not in
the following way. Let points v1 and v2 be the two endpoints of a fragment and
their nearest points on the reference ridge be points p1 and p2 respectively. l1 is
the length of the path p between p1 and p2 on the reference ridge, l2 is the length
of the fragment, and di is the distance between vi and pi , i = 1, 2, and t1 and t2
are thresholds. We will consider this fragment as an artifact instead of a ridge
section,
thus deleting this fragment if one of the following conditions is true:
l1 l2
max l2 , l1 > t1 , or d1 + d2 > t2 . The former states that the lengths l1 and l2
differ too much and the latter states that the endpoints of the fragment are too
far away from the closest ridge. Otherwise, we will keep the fragment, and delete
the pixels between p1 and p2 along p. After dealing with all the fragments in the
current image, the reference ridge becomes the form shown in Fig. 2(b). It can
be readily found from Fig. 2(b) that the original entire reference ridge has been
broken into several segments, and it is noteworthy that the two endpoints of two
different fragments in the current image should be linked only if their nearest
points on the reference ridge are the endpoints of the same segment. Therefore
we can merge the ridge parts in the current image together using this criterion.
3.3 Linking Method
Based on the aforementioned criterion, we can easily select the two endpoints
which should be linked and then connect them by a curve parallel to their adja-
cent ridge. First of all, we find the closest ridge for each endpoint. As illustrated
224 C. Liu et al.
Fig. 2. Determining which two endpoints should be linked. (a) The previous ridge
image and the current ridge image. (b) The endpoints in current image whose nearest
points on the reference ridge are the two endpoints of a segment should be linked.
Fig. 3. Linking the two endpoints which should be linked.
Fig. 4. Getting the entire ridge. (a) The current ridge image. (b) The previous ridge
image. (c) An intermediate result. (d) The final result
in Fig. 3, given an endpoint pi , a line which passes through pi and is perpen-

dicular to the orientation at pi meets the upper closest ridge at mi , and the
lower closest ridge at ni . The ridge where mi or ni lies are the closest ridge of
pi . Providing that there exist two endpoints pi and pj which ought to be linked,
they can be linked on condition that they have at least one same closest ridge.
As depicted in Fig. 3, pi and pj have two same closest ridges. We first generate
a curve Q1 which is as parallel to curve mi mj as possible to connect pi and pj ,
and then check if this curve intersects the existing ridges at any places other
than pi and pj . If not, Q1 can be used to link pi and pj together, so if pi and
pj only have one same closest ridge, we will connect pi and pj by Q1 and then
begin to check other endpoint pairs. If pi and pj have another same closest ridge,
as shown in Fig. 3, we will use ni nj to generate another as-parallel-as-possible
curve Q2 in the same way, and evaluate Q1 and Q2 by the following equation:
min(mi pi , mj pj ) min(ni pi , nj pj )
H1 = 1 − , H2 = 1 − . (1)
max(mi pi , mj pj ) max(ni pi , nj pj )
Equation 1 reflects the similarity of the distances from pi and pj to their closest
ridge. The smaller H1 (respectively H2 ) means pi and pj are at a more similar
distance to mi mj (respectively ni nj ) than to ni nj (respectively mi mj ), so it is
more suitable to use the curve generated from mi mj (respectively ni nj ) to link
pi and pj . Eventually the curve which has smaller H -value will be chosen to
connect pi and pj together. We repeatedly use this algorithm to process all pairs
of endpoints which should be linked until no pair of endpoints should be linked
in the current image.
3.4 Ridge Propagation Result

We show a result of ridge propagation. We set t1 = 4 and t2 = 10 in our
experiments. The current ridge image (Fig. 4(a)) consists of many fragments.
The reference ridge image, i.e. the previous ridge image, is shown in Fig. 4(b).
These are used by the propagation method in Sect. 3.2 to produce the result in
Fig. 4(c). The endpoint pairs in Fig. 4(a) corresponding to the two endpoints of a
segment in Fig. 4(c) are connected by the method in Sect. 3.3. The final result is
displayed in Fig. 4(d), in which all the fragments in Fig. 4(a) have been correctly
merged together, producing an entire film ridge in the current image.
Once the ridges are recovered, we use the ink projection method in [8,10]
to obtain a flattened image, which contains all the frames of the footage (see
Fig. 5(top)). Since the ridges contain the footage frames, when performing pro-
jection only a thin layer close to the ridges need to be considered (the maximum
displacement in either direction is set to 2 pixels).
4 Extracting Frames
Once the reconstructed image containing the film strip has been obtained, the
final task is to extract the individual frames so that they can be reassembled as a
226 C. Liu et al.
Fig. 5. Straightening a section of reconstructed film strip. (a) initial reconstructed

film strip showing sinusoidal deformation, (b) sinusoid fitted to the boundary curves
(shown in green) and sampled normals (shown in red), (c) film strip after warping and
cropping. (Color figure online)
video. As can be seen in Fig. 5a there is an approximately sinusoidal oscillation,

the period of which increases by a factor of about 2.7 from the beginning to
the end of the reconstructed film strip. This oscillation is due to the film not
being aligned perfectly vertically when placed in the X-ray scanner, while the
increasing period is a consequence of the spiral roll of the film. To simplify
subsequent processing, the film strip is straightened to correct for the oscillation.
First, the boundaries of the central part of the film strip that contains the
images is extracted. Since the area of the image outside this area is relatively
bright a simple approach is sufficient. The height H of a single frame1 is manually
determined, and then for each column in the image the range [ic , ic + H] that
i+H
minimises the summed intensities ic = argminr r=i I[r][c] is considered as the
frame. The set of frame boundaries ic ; c = 1 . . . n provides the upper boundary for
the film strip. Since each point along the boundary is estimated independently,
they can be noisy, so robust smoothing is applied to the boundary which down-
weights the effect of outliers.
The boundary is modelled as piecewise sinusoidal (with a fixed frequency
per section). Since a sinusoidal function is non-linear, fitting to data tends to
be unreliable unless a good initialisation is provided. The amplitude is already
known from the height H of a single frame, and the mean height of the bound-
1
Height H refers to the height of the frame which is in the film strip, and is therefore
rotated 90◦ to the normal viewing position.
(a) (b)
(c) (d)
(e) (f)
Fig. 6. Examples of problems for segmentation of film strip into frames. (a) nonlin-
ear stretching of frame, (b) nonlinear compression of frame, (c) first frame smaller
than second frame, (d) two frames with weak frame boundary, (e) first frame contains
reconstruction artifact, (f) bright patches.
ary curve provides the offset. The frequency is determined by computing the
autocorrelation of the sampled boundary function, and finding the lag that pro-
duces the first peak in the autocorrelation function. Finally, the first peak in the
boundary function is detected, providing the phase.
This procedure is applied to small sections (5000 pixels long) of the film strip
to estimate the upper boundary curve. The lower boundary is taken as the same
curve vertically shifted by H. Figure 5b shows the fitted curves for a part of a
section. A sparse set of corresponding points are found in the upper and lower
curves by computing the normals from the sinusoidal function, and projecting
them from the upper curves towards the lower curves. A thin plate spline warp
is applied to the image using the endpoints of the normals in the source image
which are mapped to a set of vertically aligned normals, separated by the arc-
length along the sinusoid, in the target image. Figure 5c shows the warped image
(after cropping) which has effectively straightened that section of the film strip.
228 C. Liu et al.
The results for the individual sections are concatenated back to form a single
150336 × 281 pixel image.
Once the film strip has been straightened and cropped it should be straight-
forward to split the image into frames. However, in practice this is problematic
due to several factors. First the process of recovering the film strip from the XMT
image volume introduces nonlinear deformations in the geometry. Examples can
be seen clearly in Fig. 6, in which parts of the image are stretched (see Fig. 6a)
or compressed (see Fig. 6b). As a consequence, the frames have been sheared, so
that the vertical frame border has become diagonal. Figure 6c shows a pair of
frames in which there has been a 10 % change in width. In addition, the image
quality is poor. Not only are the images noisy and low resolution (213 × 281
pixels), but there are several types of artifacts:
– Due to the partial volume effect banding is visible in many of the frames in
Fig. 6. The curved film surface results in the curvature of the bands.
– Figure 6e shows an example of image reconstruction errors (in the face).
– At the edges of frames there are sometimes some bright patches, e.g. Fig. 6a
and f. The film tends to be a little wrinkled at the edges, and consequently
these are also manifestations of the partial volume effect.
Finally, the boundary between frames is often not evident. For instance in Fig. 6d
and e most of the boundary does not contain any significant vertical discontinuity
in intensity or texture.
Fig. 7. Extracting frames. The green lines indicate estimated regular frame boundaries,
while the red lines show the keyframes that were determined to start a new shot. The
extracted frames are shown below after resizing to a common width (Color figure online)
Fig. 8. An X-ray image of a section of ‘The Morecambe and Wise Show’.
To split the film we provide the width W of the first frame F1 in the film
strip. All other frames will be resized to this width. The next frame F2 is assumed
to follow, but may have width W ± w. The set of possible frames F2 are com-
pared against F1 and the one maximising a normalised cross-correlation score
is selected. Once frame Fi is found it is used to estimate the width of the next
frame Fi+1 . In addition, a threshold on the normalised cross-correlation score
is used to check that the match is sufficiently good. Otherwise there may be
substantial distortion between frames Fi and Fi+1 . Alternatively, there could be
substantial change in appearance between the frames, either from a fast moving
object, sudden changes in lighting, or a cut between scenes. This would make
the estimation of the width of Fi+1 unreliable, and so it is re-estimated by com-
paring it against the currently undetermined Fi+2 . Both the widths of Fi+1 and
Fi+2 are optimised using normalised cross-correlation. This process is demon-
strated in Fig. 7. Note the width of the second frame of the woman shot has been
correctly estimated although it is more than 10 % narrower than the keyframe.

In this paper we have shown the feasibility of a primarily automatic method for
recovering the contents of lost films from source acetate rolls by performing a
virtual unrolling of a 3D tomographic scan of the film. Our approach involved
first generating the film reel by extracting lines (intensity ridges) along the film
layers, connecting disconnected lines, aligning the lines from across the X-ray
scans, and finally projecting densities along normals to the lines to measure the
response of the silver content in the film. Finally, frames were extracted from the
film reel, and reassembled into a video. Difficulties were encountered in the final
stage due to stretching artifacts in the alignment stage of the virtual unrolling.
230 C. Liu et al.
Nevertheless, reasonable results were obtained, which could be improved in the

future by better flattening methods.
The success of recovering this short section of film from a tomographic scan
gives hope for recovery of large, damaged film reels. Unfortunately, the X-ray
dose requirement increases dramatically with increasing size, so upscaling is
problematic and even impossible to the degree that would be necessary for recov-
ery of a 30 cm diameter film reel. Not only would it be necessary to increase the
detector array size to around 20,000 pixels across, but the number of recorded
photons per pixel would also have to increase by thousands. This would require
an, as yet unconceived, detector with a massive dynamic range. Furthermore,
acquiring such a photon count in a reasonable amount of time would require
such an intense X-ray beam that heating would be a problem, bearing in mind
that any dimensional change should be much less than 1 in 20,000. Conceiv-
ably, the smallest 8 mm reels might be able to be scanned in one go, whereas
larger film reels would need to be cut into smaller pieces. An alternative might
be tomosynthesis, which would avoid the necessity of an X-ray path that passes
diametrically through the film; however it is unlikely that this could produce a
reconstructed image of sufficient quality.
Returning to the rediscovered episode of ‘The Morecambe and Wise Show’, its
rapid degradation, along with the above limitations, meant that the only feasible
solution was to cut the film into smaller sections which could then be scanned
using the MuCAT 2 X-ray microtomography scanner at QMUL. Figure 8 shows
a single image from one such section. The quality is significantly poorer than the
Snow White scan, as it contains many holes and also some local deformations.
We are just starting to work on this more challenging data, and we believe that
our processing pipeline can cope since it was adapted from the pipeline developed
for damaged historical scrolled parchments [8,10], and that these documents also
contained many artifacts such as holes, delaminations, geometric distortions, etc.
Acknowledgments. We thank Paul Vanezis (BBC) and Ulrich Reudel (BFI).
References
1. Binder, M.: A Light Affliction: A History of Film Preservation and Restoration
(2015). Lulu.com
2. Davis, G.R., Evershed, A.N.Z., Mills, D.: Quantitative high contrast X-ray micro-
tomography for dental research. J. Dent. 41(5), 475–482 (2013)
3. Davis, G.R., Mills, D.: Brute force absorption contrast microtomography. In: Pro-
ceedings of SPIE: Developments in X-Ray Tomography IX, vol. 9212 (2014)
4. Macchia, A., et al.: Characterization and weathering of motion-picture films with
support of cellulose nitrate, cellulose acetate and polyester. Procedia Chem. 8,
175–184 (2013)
5. Freeth, T., et al.: Decoding the ancient Greek astronomical calculator known as
the Antikythera Mechanism. Nature 444, 587–591 (2006)
6. Enticknap, L., Restoration, F.: The Culture and Science of Audiovisual Heritage.
Palgrave Macmillan, New York (2013)
7. Kokaram, A.C.: Motion Picture Restoration: Digital Algorithms for Artefact Sup-
pression in Degraded Motion Picture Film and Video. Springer, London (2013)
8. Liu, C., Rosin, P.L., Lai, Y.-K., Hu, W.: Robust segmentation of historical parch-
ment XMT images for virtual unrolling. In: Digital Heritage, pp. 11–18 (2015)
9. Mills, D., Curtis, A., Davis, G., Rosin, P.L., Lai, Y.-K.: Apocalypto: revealing the
Bressingham roll. J. Pap. Conserv. 15(3), 14–19 (2014)
10. Samko, O., Lai, Y.-K., Marshall, D., Rosin, P.L.: Virtual unrolling and information
recovery from scanned scrolled historical documents. Pattern Recogn. 47(1), 248–
259 (2014)
11. Steger, C.: An unbiased detector of curvilinear structures. IEEE Trans. Pattern
Anal. Mach. Intell. 20(2), 113–125 (1998)
Multi-spectral Imaging System
(IWN) for the Digitization and Investigation
of Cultural Heritage
Ibrahim El-Rifai1(&), Hend Mahgoub2, and Ari Ide-Ektessabi1

1
Advanced Imaging Technology Lab, Graduate School of Engineering,
Kyoto University, Kyoto, Japan
2
Institute for Sustainable Heritage, University College London, London, UK
[email protected]
Abstract. This research focuses on the digitization and investigation of cultural

heritage liaised with the practical requirements of conservators and museum
curators. Different types of information are extracted about the physical char-
acteristics of the artifacts, pigments preliminary identification and pigments
distribution in addition to the colorimetric information. In this regard, a
multi-spectral digitization system – named as “iwn” was developed to collect the
required information from the cultural heritage objects. The system is portable,
customizable, easy to use, in-situ, non-invasive and relatively not expensive.
This paper will describe the specifications of the system showing its functions
and capabilities through few case studies.
Keywords: Digitization Multi-spectral imaging Pigment identification and

distribution Spectral reflectance reconstruction
1 Introduction
The digital era introduced several e-preservation and investigation systems for heritage
objects, such systems have been developed with variety of setups and features, ranged
from large systems [1–4] to portable and small ones [5–9]. The cost of such systems
varies significantly according to the functions and requirements.
This research aims to introduce a multi-spectral imaging system which is portable,
in-situ, non-invasive, relatively not expensive. It is designed to extract sufficient
information from the surface and also from other layers of artworks through imaging in
the visible, very near infrared (IR) and near ultraviolet (UV) regions to enrich its
documentation and conservation plans.
The system also is using a method for spectral reflectance reconstruction named as
Adaptive pseudo inverse (API) [10, 11] to improve the accuracy of the colorimetric
measurements.

DOI: 10.1007/978-3-319-48496-9_19
Multi-spectral Imaging System (IWN) 233
2 Experimental MSI System (IWN)
The multispectral imaging (MSI) system will be referred to as “iwn” throughout the
following text which is an ancient Egyptian term for color [12, 13].
IWN (Fig. 1) consists of two motorized stages for the X and Y directions, two filter
wheels facing each other, and various light sources with an ability to use different
imaging sensors. Ball screw motion is being used for the X and Y movement controlled
by stepper motors that gives movement of 5 mm translation for a full revolution about
25 µ per a single step with an option to go down to 1.5 µ movement in the
micro-stepping mode.
Fig. 1. IWN System CAD drawing (Left) and its components (Right). (Color figure online)
2.1 System Specifications

2.1.1 Lights
Different light sources are used to cover the required regions; six Ultraviolet
(UV) lamps (Mini Blacklight - 6 W UV-A), two infrared (IR) Led light (850 nm)
sources and two Visible (VIS) led Light (5000k) sources.
2.1.2 Filter Wheels

Two filters wheels are used to host different combinations of filters. The first wheel is
designed to host 7 filters (Fujifilm BandPass) covering the visible spectrum in; 420,
450, 500, 530, 550, 600 and 640 nm which can be replaced with other filters as
required. The second wheel hosts 3 filters; IR reflected filter (B + W 093), UV reflected
filter (B + W 403) and IR/UV block filter - hot mirror (B + W 486). The wheel is used
to pass either infrared or ultraviolet only or block both to pass only the visible radiation.
The second wheel can be replaced by another filter wheel that can host either another 7
filters in the visible range to achieve 14 different filters combinations or a range of IR
filters from 700–960 nm as needed.
234 I. El-Rifai et al.
2.1.3 Imaging Sensors

IWN can be used with either a monochromatic or DSLR cameras. The monochromatic
sensor is an area camera Artray ArtCam-150PIII with c-mount to f-mount adapter to
allow the use of Nikon 35 mm f/2D lens.
Many modified DSLR cameras can be installed with the system which are sensitive
to the UVA and NIR radiation. A modified Fuji FinePix S5 camera with a sensitivity
range from 380–1000 nm was used in some cases.
2.1.4 Software
A software was developed to control IWN system in two modes; full control mode or
DSLR mode. The full control mode controls the whole system (X and Y motion, filter
wheels and the different types of lights) in addition to the monochromatic camera.
The DSLR mode also controls all the system parameters except the DSLR camera
parameters which are controlled by the manufacturer software.
2.2 Object Specifications

The monochromatic setup of IWN allows capturing an object of 2 × 1.5 cm and a
maximum of 14.5 × 10.5 cm in size on a distance of 12–80 cm from the object. Wider
objects can be captured using the
motorized system that moves 60 cm
in the X and Y if needed. However,
these specifications can be increased
with the use of DSLR camera or the
use of different lenses.
The system can be installed in
different ways to allow for horizon-
tal or vertical imaging. It also can be
attached to linear guide rails to give
the ability of capturing longer
objects like papyrus and manuscripts Fig. 2. IWN on linear guide rails for investigation
(Fig. 2). of long objects (Color figure online)
2.3 System Calibration

Different calibration targets (i.e. Labsphere white reference and X-Rite color checker
passport) are used to improve the accuracy and reproducibility of the system. In
addition to the collection of dark, bias frames and flat field frames. An image reduction
process also is used to remove the noise from the electronics or from the system optics.
The removal of the dust from the CCD and the calculations of system gain and signal to
noise ratio are performed regularly to insure the stability of the system.
3 Case Studies
3.1 Spectral Reflectance Reconstruction
A test painting was captured using IWN to reconstruct the spectral reflectance of the
different colors from the multi-spectral data. An area (12 × 10 cm) was selected
(Fig. 3) for the experiment.
Fig. 3. Oil Painting (test painting). Black square represents the selected area for investigation
(Left), Monochromatic image of the selected area showing the 21 selected positions of the
training set (Middle) and the colorimetric measurements of these positions. (Color figure online)
In this experiment, 7 filters were used in the range of 420–700 nm (Fujifilm

BPB42, BPB45, BPB50, BPB55, BPB60, SC64, SC70) accompanied by UV/IR block
filter. The acquisition was conducted in a dark room. 21 positions (Fig. 3) were
selected for the training dataset. The reflectance of these positions was measured by
X-Rite eXact colorimeter and re-sampled in the range of 400–700 nm with 10 nm
interval. Spectral data from the positions were collected in two modes; one pixel and
region of interest (ROI) with an average area of 4 mm similar to the size of the
colorimeter scan area.
All the measurements were used to reconstruct spectral reflectance using two
estimation methods for comparison; the pseudo inverse (PI) [14] and the adaptive
pseudo inverse method (API) [10]. The results of the comparison (Table 1) were
assessed by the mean squared error equation for spectral error between the actual and
the estimated spectral reflectance and the equation of ΔE2000 obtained from the for-
mula of CIEDE2000 [15, 16] under the CIE D65 as a standard illuminant for colori-
metric error.
The proposed method API is showing noticeable improvement over the PI method
in both modes (one pixel and ROI).
It should be considered that the data was not acquired from a standard color chart
with a homogeneous color patches so it is expected to find a mixture of pigments or at
least average of different intensity levels of the same pigment, which is reflected in the
values of the standard deviation in spectral and colorimetric errors. The new method is
showing a potential to be used to improve the accuracy of color reconstruction but
further work is needed which is the focus of our further research.
Table 1. Spectral and colormetric errors for the 21 positions of two spectral reflectance
construction methods in two modes; one pixel and region of interest (4 mm) – ROI. The table is
showing the spectral RMS and colorimetric errors including the mean, standard deviation and
maximum error of the tested methods for the 21 positions dataset.
21 positions One pixel ROI
API PI Diff. API PI Diff.
RMSE Mean 0.0026 0.0043 −0.0017 0.0019 0.0039 −0.0019
STD 0.0032 0.0027 0.0005 0.0037 0.0041 −0.0004
Max 0.0114 0.0104 0.0011 0.0157 0.0184 −0.0027
ΔE00 under D65 Mean 0.0844 0.1455 −0.0611 0.0507 0.1250 −0.07438
STD 0.0909 0.0988 −0.0079 0.0864 0.1431 −0.0567
Max 0.2537 0.3905 −0.1368 0.3446 0.6370 −0.2924
3.2 Oil Painting Investigation

Multispectral imaging has been used intensively in the investigation of paintings [17–
22] due to its effectiveness and valuable results. In this section a painting from the
storage of the Egyptian geographic society will be presented as a case study for
investigating paintings using IWN (Fig. 4). The painting is suffering from many
deterioration aspects and was found without frame or information about the portrayed
figure. Comparing the painting with other portraits in the geographic society, it was
suggested that this frameless portrait belongs to one of the famous figures of the
geographic society with no more information.
Fig. 4. Unknown oil painting from the Egyptian geographic society (Color figure online)
The painting (verso and recto) was digitized in the visible, near IR and UV ranges
followed by various post-processes to explore any features or hidden remarks that
could lead to know the depicted person in addition to the documentation of the painting
condition.
The initial inspection of the images revealed a change in the composition in the left
arm of the man, which was assumed positioned behind his back. Through multispectral
imaging, it is now clear that the artist painted the left arm in the original composition
(Fig. 5), then removed it intentionally. This worked as a guide to search for a pioneer
who had a problem with his left arm which did not take too long to be revealed “Dr.
Livingstone”.
David Livingstone (19 Mar. 1813–1

May 1873) is a famous European
explorer who was seriously wounded
and had a disability of his arm after
being attacked by a lion in one of his
expeditions [23].
It is believed that the portrait was
painted from an engraving made by D.
J. Pound on 1857 from a photograph by Fig. 5. Details of the painting (Visible and IR
J.J. Edwin Mayall [24]. images of the left arm) (Color figure online)
A previous conservation also is
observed around the head (Fig. 4.d)
which shows the different degradation rate of pigments between the original and the
retouched.
Unfortunately, there is no signature or date for this painting but it is estimated to be
painted at the end of the 19th century.
3.3 Pigments Preliminary Identification and Distribution

MSI proved to be useful for pigments
preliminary identification [25–31]. 9 pig-
ments were selected to test the ability of
IWN to identify and differentiate between
the different pigments; Lapis Lazuli,
Ultramarine, Egyptian Blue, Azurite,
Smalt, Malachite, Verdigris, Viridian,
Cadmium Red and Cadmium Yellow (L.
Cornelissen & son [32]).
Pigments reaction were investigated
under several multi-spectral imaging
modes of acquisition such as: Visible,
Visible induced IR (VIL / IRF), UV
reflected, UV fluorescence, and IR reflec-
Fig. 6. Pigments reaction under multispec-
ted in addition to producing UV and IR
tral acquisition. (Color figure online)
false color images.
Observations:
Blue Pigments
Egyptian blue has a unique infrared luminescence property when excited with the
visible light and captured using infrared filter. Thus, it is easily identified among other
pigments [29, 30]. Azurite also reacted similarly to Egyptian blue under the different
modes except for the infrared luminescence mode.
Smalt and Azurite can be differentiated by observing the IR false color image as smalt
appears red while Azurite is purple.
Ultramarine and Lapis Lazuli have similar responses under MSI technique which is
expected as both are chemically similar [33]. Both can be differentiated to Azurite by
IR false color image.
Green pigments
Green pigments gave different responses as seen in Fig. 6. Malachite turns yellow in
the UV false color and blue in the IR false color. On the other hand, Viridian turns
yellowish in the UV false color while Verdigris gives greenish color. Also, Viridian
can be identified by the pink or reddish color in the IR false color while Verdigris
turns dark.
Cadmium pigments: (Red and Yellow)
Cadmium pigments can be identified by the faint glow or fluorescence in the VIL
mode, Moreover, cadmium yellow gives red color in the UV false color while cad-
mium red becomes dark. Cadmium red also gives orange-yellowish color in the IR
false color.
4 Conclusions
This research presented a multispectral imaging system (IWN) which is portable,

customizable, in-situ, non-invasive and relatively not expensive. The system is working
in the range of 380–1000 nm. It is developed with the practical requirements of
museum curators and conservators in mind which they need a fast tool to digitize and
investigate cultural heritage objects. The system has the ability to acquire color
information with the potential to improve the accuracy of the reconstruction of colors
through the use of modified spectral reflectance reconstruction method (Adaptive
Pseudo-inverse - API). The system is also capable of the digitization and investigation
in the visible, near infrared and near ultraviolet regions which are imaging techniques
proved its usefulness/utility in the investigation of different artifacts such as oil
paintings in addition to the potential of preliminary identification of early pigments.
Acknowledgement. This research is partially funded and supported by Prof. Ide-Ektessabi

Laboratory, Advanced Imaging Technology, Graduate School of Engineering, Kyoto University
as part of a PhD research.
References
1. Martinez, K., Cupitt, J., Saunders, D., Pillay, R.: Ten years of art imaging research. Proc.
IEEE 90, 28–41 (2002)
2. Ribés, A., Schmitt, F., Pillay, R., Lahanier, C.: Calibration and spectral reconstruction for
cristatel: an art painting multispectral acquisition system. J. Imaging Sci. Technol. 49, 563–
573 (2005)
3. Christens-Barry, W.A., Boydston, K., France, F.G., Knox, K.T., Easton Jr., R.L., Toth, M.
B.: Camera system for multispectral imaging of documents. In: Proceedings of SPIE 7249,
Sensors, Cameras, and Systems for Industrial/Scientific Applications X, 724908, 27 January
2009. doi:10.1117/12.815374
4. Easton Jr., R.L., Knox, K.T., Christens-Barry, W.A.: Multispectral imaging of the
Archimedes palimpsest. In: Applied Imagery Pattern Recognition Workshop Proceedings,
pp. 15–17 (2003)
5. Costas, B., et al.: A novel hyper-spectral imaging apparatus for the non-destructive analysis
of objects of artistic and historic value. J. Cult. Heritage 4, 330s–337s (2003)
6. Novati, G., Pellegri, P., Schettini, R.: An affordable multispectral imaging system for the
digital museum. Int. J. Digit. Libr. 5, 167–178 (2005)
7. Pelagotti, A., Del Mastio, A., De Rosa, A., Piva, A.: Multispectral imaging of paintings.
IEEE Sign. Process. Mag. 25, 27–36 (2008)
8. Berns, R.S., Taplin, L.A., Urban, P., Zhao, Y.: Spectral color reproduction of paintings. In:
Proceedings of the Fourth European Conference on Color in Graphics, Imaging and Vision,
Barcelona, Spain, pp. 484–488 (2013)
9. Bearman, G.H., Spiro, S.I.: Archaeological applications of advanced imaging techniques.
Biblical Archaeologist 59(1), 56–66 (1996)
10. El-Rifai, I., Mahgoub, H., Magdy, M., Toque, J.A., Ide-Ektessabi, A.: Enhanced spectral
reflectance reconstruction using pseudo-inverse estimation method. Int. J. Image Process.
(IJIP) 7(3), 278–285 (2013)
11. Ohno, Y.: CIE fundamentals for color measurements. In: International Conference on Digital
Printing Technology, NIP 2016, Vancouver, Canada, pp. 540–545 (2016)
12. Faulkner, R.O.: A Concise Dictionary of Middle Egyptian Oxford, p. 13. Oxford University
Press, London (1991)
13. Gardiner, A.: Egyptian Grammar: Being an Introduction to the Study of Hieroglyphs, 3rd
edn, p. 552. Oxford University Press, London (1966)
14. Barata, J., Hussein, M.: The moore-penrose pseudoinverse: a tutorial review of the theory.
Braz. J. Phys. 42(1–2), 146–165 (2012)
15. ASTM International: Standard Practice for Computing the Colors of Objects by Using the
CIE System, ASTM Standard E308-08, West Conshohocken, United States (2008). www.
astm.org
16. Luo, M.R., Cui, G., Rigg, B.: The development of the CIE 2000 colour-difference formula:
CIEDE2000. Color Res. Appl. 26(5), 340–350 (2001). (Special Issue on Color Difference)
17. El-Rifai, I., Ezzat, H., Mahgoub, H., Bebars, Y., Ide-Ektessabi, A.: Artwork digitization and
investigation a case study of the loom weaver, oil painting by Hosni El-Bannani. Int. Sci.
J. Mediterr. Archaeol. Archaeometry 13(2), 21–29 (2013). Greece
18. René de la Rie, E.: Fluorescence of paint and varnish layers (part I). Stud. Conserv. 27(1), 1–
7 (1982). doi:10.2307/1505977
19. René de la Rie, E.: Fluorescence of paint and varnish layers (part III). Stud. Conserv. 27(3),
102–108 (1982)
20. Walmsley, E., Metzger, C., Delaney, J.K., Fletcher, C.: Improved visualization of under
drawings with solid-state detectors operating in the infrared. Stud. Conserv. 39, 217–231
(1994)
21. Aldrovandi, A., Buzzegoli, E., Keller, A., Kunzelman, D.: Investigation of painted surfaces
with a reflected UV false color technique. In: 8th International Conference on Non
Destructive Investigations and Micronalysis for the Diagnostics and Conservation of the
Cultural and Environmental Heritage, Lecce, art 2005, Italy (2005)
22. Pelagotti, A., Del Mastio, A., De Rosa, A., Piva, A.: Multispectral imaging of paintings.
IEEE Sign. Process. Mag. 25, 27–36 (2008)
23. Livingstone, D.: Encyclopaedia Britannica Online (2016). www.britannica.com/biography/

David-Livingstone. Accessed June 2016
24. Pound, D.J.: The Drawing-Room of Eminent Personages, vol. 2. Mayall, London (1857).
John Jabez Edwin Paisley. npg.org.uk/collections/search/portrait/mw40836/David-
Livingstone?
25. Toque, J.A., Sakatoku, Y., Ide-Ektessabi, A.: Pigment identification by analytical imaging
using multispectral images. In: 2009 16th IEEE International Conference on Image
Processing (ICIP), Cairo, pp. 2861–2864 (2009)
26. Delaney, J.K., Walmsley, E., Berrie, B.H., Fletcher, C.F.: Multispectral Imaging of Paintings
in the Infrared to Detect and Map Blue Pigments. National Academy of Sciences. (Sackler
NAS Colloquium) Scientific Examination of Art: Modern Techniques in Conservation and
Analysis. The National Academies Press, Washington, DC (2005). doi:10.17226/11413
27. Delaney, J.K., Walmsey, E., Berrie, B.H., Fletcher, C.F.: Multispectral imaging of paintings
in the infared to detect and map blue pigments. In: Proceedings of the National Academy of
Sciences, pp. 120–136 (2003)
28. Smith, G.D., et al.: Inexpensive, near-infrared imaging of artwork using a night-vision
webcam for chemistry-of-art courses. J. Chem. Educ. 86(12), December 2009. Verri, “The
spatially resolved characterization of Egyptian blue.”
29. Verri, G.: The spatially resolved characterization of Egyptian blue, Han blue and Han purple
by photo-induced luminescence digital imaging. Anal. Bioanal. Chem. 394(4), 1011–1021
(2009). doi:10.1007/s00216-009-2693-0
30. Accorsi, G., Verri, G., Bolognesi, M., Armaroli, N., Clementi, C., Miliani, C., Romani, A.:
The exceptional near-infrared luminescence properties of cuprorivaite (Egyptian blue).
Chem. Commun. 23, 3392–3394 (2009). doi:10.1039/B902563D
31. Cosentino, A.: Identification of pigments by multispectral imaging a flowchart method.
Heritage Sci. 2(8) (2014). doi:10.1186/2050-7445-2-8
32. L. Cornelissen & Son, London: www.cornelissen.com
33. Sackler Colloquia: Scientific Examination of Art: Modern Techniques in Conservation and
Analysis, Arthur M. Sackler Colloquia of the National Academy of Sciences, National
Academies Press, Washington, D.C. (2003)
Diagnostic Activities for the Planned and Preventive
Conservation of Mosaic Pavements: The Case Study
of the Triclinium of the Villa Romana del Casale (Sicily)
Antonella Versaci1, Alessio Cardaci2 ✉ , and Luca R. Fauzia1

( )
1
Faculty of Engineering and Architecture, University of Enna KORE, Enna, Italy
{antonella.versaci,luca.fauzia}@unikore.it
2
School of Engineering, University of Bergamo, Bergamo, Italy
[email protected]
Abstract. The Villa Romana del Casale at Piazza Armerina, Sicily is known for
the richness of the mosaic pavements that decorate almost every room. They are
the finest mosaics from the Roman world and, even because of their exceptional
extent, the Villa was declared a UNESCO World Heritage site in 1997. Their
in situ conservation advocated by Cesare Brandi in the late 1950s, an exception
to the typical treatment of excavated mosaics of the time, requires a regular
monitoring of the physical condition and the establishment of mitigation strat‐
egies, however, difficult to implement especially for economic and technical
reasons. In this sense, this paper intends to propose an innovative and user-
friendly procedure based on laser scanning and thermo-hygrometric investiga‐
tions able to assess the tessellatum status through time and to evaluate the main‐
tenance work’s efficiency, which could be included in a long-term and sustainable
approach to preserving our ancient mosaic heritage.
Keywords: Mosaics · Preventive conservation · Diagnostics · Piazza Armerina
1 Introduction
The mosaics, one of the main wealth of the Mediterranean area, represent a treasure of
extreme delicacy, often endangered by both the neglect and the deficiency of economic
resources, if not subverted by other ever-increasing man-made disasters such as wars
and terrorist attacks. The in situ conservation of mosaic decorations is at present a largely
privileged cultural choice, which, in concert with the site’s documentation and valori‐
sation, requires, since the time of excavation, to study and implement a series of coor‐
dinated and systematic acts to counter the decline of their physical and chemical state,
as well as the growing of instability phenomena. Too often, however, the carelessness
of preliminary investigations, the absence of timely and appropriate conservation/main‐
tenance measures, the inadequacy of the protection systems from the major climatic
factors, the indiscriminate influx of tourists and even the consequences of intrusions,
theft and vandalism, often favoured by the failure of tools and security/surveillance
activities, cause irreversible damages. In order to oppose the destruction of this precious

DOI: 10.1007/978-3-319-48496-9_20
242 A. Versaci et al.
heritage, it appears necessary to develop a culture of planning and protection, which

could finally supplant that of emergency and improvisation; i.e. to establish principles
able to improve an idea of well-structured preservation. Nevertheless, for this to happen,
despite the progress made by the matter for the past sixty years, it is still essential to
operate “a profound change in mentality” [1], from how to why heritage should be
preserved [2], both in the operators and administrations.
Today, conservation techniques seem having definitively abandoned the practice of
mosaics detachment long-followed and motivated by a desire to safeguard only their
image, paying little attention to their material consistency and heedless of their three-
dimensional values [3, 4]. Nonetheless, the maintaining in the original location first
demands a great ethical consciousness, as well as specific operational and management
capacities. In most cases nowadays, the mosaic is restored in situ, without altering the
final appearance of the surfaces, respecting the traces of the past and the historical/artistic
values of the good and its archaeological field. Conservation practice does not in fact
solely apply to mosaic but goes further to the analysis of its context, the building that it
covers and to which is intimately linked, the landscape and highly considers the signif‐
icance of the work itself. Major new activities are then entrusted to the restorer, now
necessarily part of a multidisciplinary team composed of engineers, chemists, physicists,
biologists, mineralogists, etc. Starting from a deep knowledge of the mosaic, its func‐
tionality, techniques of execution, original materials and those used in previous resto‐
rations and from the documentation of its state of preservation, he plans the intervention
to be executed [5]. He also studies the measures to be taken for the future protection of
the property: indirect interventions on the environment [6], monitoring and prevention
seen as “a mode that maximizes the permanence of authenticity (material, of course) of
architectural object” [7]. These devices, critical to secure a future to works that are an
expression of art, technology and building expertise and a privileged source of knowl‐
edge related to ancient civilizations, still seem hard to implement, mainly due to both
the lack of adequate resources and technological skills. They demand a full awareness
of the numerous meanings embodied in cultural heritage even in users and authorities
that should be understood as owners/managers of the properties, fully involved in their
conservation through on-going maintenance programs.
In this framework, the control activities on the mosaics of the Roman Villa in Piazza
Armerina, which this work illustrates, have been planned and implemented. The monu‐
ment includes “the largest and comprehensive set of mosaics ever discovered in a single
monument, and mosaics in a state of preservation, if not perfect, some very considerable”
[8]. It is known for the complexity of figurative themes, which are very important for
the historical and aesthetical values and a cultural testimony of late imperial Rome in
the heart of the Mediterranean (Fig. 1).
Characterized by a nearly 4,000 m2 extension, the site must deal with the complex
issues of iconography’s conservation, public enjoyment and appreciation that inevitably
play an essential role in all management decisions [9]. The case study, therefore, aims
to contribute to the preservation of a priceless heritage, through the design and execution
of simple, low-cost monitoring actions, to ensure early detection of possible anomalies
and related causes, as well as of the onset of degradation phenomena and damages.
Diagnostic Activities for the Planned and Preventive Conservation 243
Fig. 1. Plan of the Villa Romana del Casale (left) and its famous mosaic known as “The Bikini
Girls”, early 4th century A.D. (right)
2 The Villa Romana del Casale and Its Mosaics
The Villa Romana del Casale is part of a wider geographical and cultural area, in which
the historical factors that have produced countless goods spread throughout the territory
are intimately connected to the magnificent characteristics of landscape, shaping today’s
morphology. Little or poorly studied by the admirers of classical antiquity and the
ancient surveyors, it is only in the second half of 18th century that, with the start of Grand
Tours, descriptions of the city and its monuments appeared. The first complete portrait
of Piazza Armerina dates back to 1761 by Arcangelo Leanti that puts the old town in a
countryside named ‘the Casale’, close to the ruins of what will prove to be an imposing
Roman villa. He depicts the surrounding lands as pleasant and delightful as later will
appear in one of the illustrations of the Voyage pittoresque by Jean Claude Richard de
Saint-Non. This image of a cosy and lush place, well away from the collective belief of
a barren and desolate inland Sicily, becomes recurring in the reports of following trav‐
ellers such as the Dutch classicist Jacques-Philippe D’Orville (1727), the great geologist
Déodat de Dolomieu (1781), the German poet Augusto Von Platen (1835). They
returned a vision of the place of exceptional interest, so to affirm that this spatial feature
has favoured in the 4th century AD the construction of the Villa del Casale, not far from
the Itinerarium Antonini, which linked Catania to Agrigento [10].
The imposing late-antique building, a “unicum in the monumental and figurative
heritage of the past” [11], was built on pre-existing structures, probably a villa rustica.
It belonged to a member of the Roman senatorial aristocracy, maybe a governor; or,
according to some scholars, it was commissioned and improved by a personality of
imperial rank, perhaps Maxentius [12]. Traces of it were found from the late 18th century
when clandestine excavations had already drawn fully from the precious heritage of
archaeological finds hidden underground. At the beginning of the 19th century, confused
and hasty researches were initiated by careless diggers or antique dealers, who caused
serious harm to mosaic floors. Sabatino Del Muto, on behalf of Roberto Fagan, British
Consul-General in Sicily, performed the first authorised excavations. However, illicit
activities continued in the following years in the absence of an effective protection

system, often only delegated to the good will of some wardens.
It was after the unification of Italy that more systematic explorations started. In 1881,
excavations were indeed conducted, under the ministerial aegis, by the engineer Luigi
Pappalardo, the inspector of monuments and excavations of antiquities of Caltanissetta
district [13]. Pappalardo will raise the issue of mosaics’ protection since the discovery
of the first fragments of ‘the twelve labours of Hercules’, which took place following
the opening of the first trench in the central part of the triclinium. It soon became clear,
in fact, the danger to which the tesserae will encounter after the removal of the thick
layer of soil that had protected them all along the centuries, both because of the weather
and the action of treasure hunters. The preservation of the mosaic will be assigned to a
shelter designed by the famous architect-restorer Piero Gazzola with “the task of trans‐
ferring a precious fragment from the past (a three-apsed hall with rich mosaic floors) to
the present times” [14]. The enormous roofing made by “wooden trusses, with terracotta
tiles, supported by massive brick pillars […] that adulterated so the appearance of the
ruin, to make it similar to a barn instead of a sumptuous room as it should be” [8], built
in the years 1941–42, was not however able to protect the mosaics by wind and heavy
weather. The fact that visitors could walk on them had caused the lifting of large parts
and the detachment of numerous tesserae. Moreover, erroneous cleanings made by
hydrochloric acid treatments had resulted in a deterioration of cohesion mortars.
New interventions were executed between 1942 and 1949 for the restoration of the
mosaic carpet. Nevertheless, it was only after the excavations carried out by Gino
Vinicio Gentili in 1950 that the architect Franco Minissi will be charged with the design
of a new protective shelter. A roof that, giving to Cesare Brandi, at that time director of
the Istituto Centrale del Restauro, should have been characterised by a non-monumental
proposal, leaving the ruins clearly visible. The design of Minissi, implemented in a first
phase in the years 1957–1960, to which followed new shelters for the basilica (1977),
the ‘corridor of the big hunt’ and the triclinium (1978), consisted of a lightweight steel
skeleton sheathed with translucent panels of plastic material (a slightly clouded Perspex
laminate, 3.2 mm thick). His aims were to protect the mosaics while retaining the
maximum light, to permit visitors to see the interior of the villa from above, without
walking upon the mosaic floors and to create new volumes suggesting the missing ones
with a modern material, while safeguarding the integrity of the ancient. The intent was
“to reform not rebuild” and to reintegrate the architectural image of the villa without
“imitating the original” [8]. For this reason was approved by Cesare Brandi who consid‐
ered it the 3D translation of that technique of tratteggio he was advocating for the inte‐
gration of losses in paintings and polychrome sculptures [17].
This solution guaranteed the maintenance of the mosaics in situ and their fruition by
millions of visitors, representing a milestone in the history of restoration and museum
display in the archaeological field because of the sensitivity and the ability to dialogue
with the landscape. However, the rapid deterioration of the materials did not respond
the expectations of the same designer, showing major limitations in the conservation of
the decorations. Compromised by serious degradation problems in the 1970s, perhaps
also due to an unexceptional realization, it was subjected to subsequent adjustment and
protection interventions in the years 1985–88, planned by Minissi. He replaced the
original Perspex ‘blinds’ for perimeter shielding of pavilions with Visarm glass sheets
12 mm thick and the original Perspex ceilings and roofs with other similar laminates in
acrylic material, as well as the opening of Vasistas windows.
In the 1990s, however, the entire mosaic decorations was in a state of extreme disre‐
pair. The deterioration of roofing and metal supports, the lack of constant maintenance,
along with a variety of events (vandalism, percolation and stagnant waters rain, moisture
in the walls and floors, thermal imbalance and humidity inside the rooms, the biode‐
teriogens organisms attack, etc.) asked for targeted interventions in the site. They then
became mandatory by the heavy rains of October 1991 that flooded the Villa extensively.
Therefore, the Regional Center for Planning and Restoration undertook new studies. It
was suggested implementing a number of measures primarily aimed at first regiment
the waters for the purpose of a hygrometer soil stabilization of the villa’s premises
because the changed moisture conditions could lead to further and more serious defor‐
mations of the mosaic floors. In particular, it was denounced the use of cement and iron
bands both for the reconstruction and consolidation of vertical structures and the resto‐
ration of the mosaics that had already caused efflorescence on the plaster and masonry,
as well as internal tensions with consequent fractures, lifting and material explosions.
It was then required the almost surgical removal of such elements, replacing them with
mortars, kicks and stones having historical and material legitimacy [16, 17].
In the 1998–2002 years, new scientific surveys were accomplished by ENEA and
the Central Institute of Restoration, highlighted several problems related to environ‐
mental comfort, attributable to incorrect maintenance solutions. The research conclu‐
sions pointed out that in the absence of opaque closures, including the use of new-
generation transparent commercially available materials, microclimatic parameters
could not be effectively controlled [18]. It then made its way the idea of a divestiture of
Minissi’s shelters, which will become effective with the guidelines for the design, which
Vittorio Sgarbi, High Commissioner for the Villa of Piazza Armerina, will in 2005
(Fig. 2).
Fig. 2. The triclinium’s shelters: Piero Gazzola’s project (left), Franco Minissi’s project (center)
and the new project by Guido Meli (left).
Despite numerous appeals by the international scientific world, the “rusty scrap
metal” will be removed [19]. In order to improve living conditions and thus promote the
conservation of the decorations, a new project designed by architect Guido Meli will be
realized in 2012. It will give up the transparency of the materials as considered unreliable
for the proper maintenance and inadequate for the use [17] and will provide for important
recovery activities of the mosaics. The new roofing system will be made by a steel frame
and laminated wood, with cladding made of panels in polystyrene layers, stucco and
aluminium plastered outside and ventilated copper roofs. It will concern the entire
monumental site except, even if foreseen by the program, the frigidarium and the tricli‐
nium; a structure, this latter, which lies today almost abandoned.
3 The Triclinium: Reasons for a Case Study
Minissi’s shelters that still remains on the Triclinium, causes considerable problems that
are accelerating degradation processes already in place. While waiting for the construc‐
tion of the new structure, this large three-apsed hall has been forbidden to visitors, both
due to the precarious condition of the path along the wall tops and the fragile state of
conservation of mosaics. The room consists of a square central atrium of about 12 m per
side, with an inlet at west punctuated by two columns, which leads to the ovoid peristyle.
On the other three sides are three semi-circular apses with a radius of almost 4 m, each
one preceded by a threshold were the traces of two pillars are still visible.
This environment, together with the adjacent peristyle (xystus), was erected in a later
stage than the rest of the villa, maybe in Theodosian era [20]. This is evidenced by
stylistic and iconological assessments of mosaics and by some findings of recent exca‐
vations, as well as by the presence of thicker walls (perhaps made in this way to deal
with the structural problems that in the basilica and other settings imposed the subse‐
quent construction of buttresses). This part of the Villa was supposed to host banquets
offered by the owner to his guests, who were welcomed in a gorgeous environment suited
to the fun and carefree, with mosaics that often refer to the cult of Dionysus. The mosaic
of the large central atrium represents ‘the twelve labours of Hercules’ and develops along
five bands parallel to the entrance. In the scene, unfortunately, today very deficient in
the west part, the ninth labour - the Girdle of Hippolyte - is missing [10]. In the mosaic
bands used as thresholds to the three exedrae, the art historian Salvatore Settis recog‐
nized a thread around the theme of metamorphosis that makes immortal: Daphne and
Cyparissus on the south side, Andromeda and Endymion on the east side, perhaps
Ampelos and Kissos on the north side [11]. The mosaic on the left of the exedra is the
apotheosis of Hercules; the central apse, the only one where there are no gaps and which
can still be appreciated in its entirety, shows anguiped giants pierced by the arrows of
Hercules. Finally, the right exedra displays the Ambrosia nymph rescued by Dionysus
from the assault of King Lycurgus.
The metal and glass structure, which now covers this environment, makes the reading
of the mosaics extremely difficult, because of the shadows cast on all surfaces. It also
causes the greenhouse effect that exposes the mosaic pavements to both high summer
heat and low winter temperatures.
The lack of adequate air conditioning system does not, therefore, protect the mosaics
from both the risks of a high thermal stress and the consequences of the condensation
that arises on the surfaces in conditions, alas, frequently, in the cold months. The shelter
also presents several shortcomings, opening access routes to the birds ‘living’ undis‐
turbed this environment, tainting of guano the mosaics and significantly contributing to
their biodeterioration. In addition, the roofing structure and the runway are highly
oxidized and the rainwater disposal plant is ineffective because one of the drainpipes is
broken and another has been removed.
This precarious situation that has continued for several years, now strongly contrasts
with the overall state of conservation of the site. Although lacking in a maintenance
program which still needs to be approved due to inexplicable bureaucratic reasons (it
has to be remembered that the site is in the UNESCO World Heritage List since 1997
and subjected to a strict management plan), the Villa seems having largely benefited
from the new roofing systems. Despite the still ongoing controversy on the overall
concept of the architectural project, it has allowed, to date, a better preservation of the
mosaic decorations and a more user-friendly fruition.
Conversely, in the triclinium, although restored in 2007 there has been a clear dete‐
rioration of the mosaics, which has sought the opening of new investigations (Fig. 3).
These have been studied in accordance with a clear desire for simplification of the
monitoring activities aimed at prevention of decay’s phenomena, also sustainable from
the economic point of view.
Fig. 3. Major vulnerabilities on the triclinium.
The goal of the experiment is not to confirm the ineffectiveness of a shelter, already
clearly compromised by tampering made during the years and by the total absence of
maintenance. It rather aims at contributing to the formulation of simple procedures for
early detection of degradation that could become part of specific protocols for preventive
and planned conservation of the whole complex.
4 Mosaics Monitoring Activities Through 3D Survey

and Thermoygrometric Analysis
The conservation of a delicate heritage as floor mosaics are, especially when preserved
on site, requires regular maintenance to reduce the impact of destructive environmental
forces. Maintenance consists of a series of operations that includes a preliminary study,
pre-consolidation work, and periodic inspections followed, if needed, by planned inter‐
ventions to protect and stabilize it. Control activities cannot be occasional and episodic
but must be done in a methodical and structured manner. They begin with collecting
data about the mosaic, its condition, and previous interventions.
The metric survey and diagnostic analysis, if regularly taken, are powerful tools for
checking goods state. The survey on the spatial surface of tessellatum, in order to identify
the changes in the geometry (new depressions, punctual elevations or areas where the
tiles rotate moving away), offers insight into the variation that has suffered over the
centuries compared to a presumed initial arrangement. It also can help in attesting the
existence of any ongoing action, responsible for alignment’s changes.
However, the diagnostic analysis is able to return an overall picture of the geophys‐
ical situation of substratum (presence and level of groundwater, soil’s imbibition, etc.)
and the microclimate just above the decorated squares (temperature, relative and abso‐
lute humidity) that can help in understanding the causes of degenerative phenomena in
place. Actually, the tesserae can break away because of the mechanical actions generated
by the movement of the mantle or even for physical and chemical reasons (the presence
of water and/or high-temperature excursions) which attack the mortar binders.
The Metric Survey and Diagnostic Tests (Microclimate and IR Analysis). The
analytical investigations carried out through metric, material, and colorimetric obser‐
vation, are able to provide clear guidance on both structural instabilities underway and
the progression/regression of degradation phenomena over time. The close range acquis‐
itions - terrestrial 3D laser scanning and photogrammetric 3D image-based reconstruc‐
tion - allow accurate and not expensive measurements, both in terms of cost and time.
The two technologies are suitable for iterative activities, thanks to their flexibility, speed,
and automation. The 3D laser scanning, in particular, is able to return in an expeditious
manner and in the form of a point clouds a very accurate three-dimensional model. On
the other hand, the photogrammetry allows creating a more simplified digital artefact
but with a very high colour rendering. The two techniques, if integrated, can deliver a
result characterised by the metric precision granted by the optically active sensors and
the high detail of the colour information instead possible with passive optical sensors.
In addition, compared to other traditional systems for precise monitoring, they are non-
contact measurement instruments that do not require targets anchored with adhesives
and/or silicone materials, whose application and removal could damage mosaics. The
surveys planning, in fact, founded on dense acquisitions at close range from the pave‐
ment, allows an appropriate resolution, which can ensure the virtual reconstruction of
the individual tiles [9, 21]. The use of a normal equipment typically allows obtaining a
polygon model with a distance between each vertex of 4–5 mm to which is projected a
texture with an interval between two sequential pixels of less than 1–2 mm.
With this in mind, the triclinium was the subject of a number of survey campaigns
that have been programmed, initially, on an annual basis: the first was undertaken in
July 2015, the second in June 2016. The acquisitions were executed from 15 stations
(the same for each survey campaign, 3 stations on the lacunae and 12 along the perimeter
of the walkway). They allowed obtaining, in addition to the plans at various altitudes
and sections/elevations, the mapping of geometric discontinuities of the surface and
height variations, with respect to a horizontal plane taken as the reference. The final
model has been produced thanks to the overrun, on the geometry acquired with 3D laser
scanning, of a single high-resolution photomapping, the result of digital photogram‐
metry activities. In both tests, shots were performed in the early morning, in diffuse light
conditions (to avoid the projection of the metal roof’s shadows on the mosaic surfaces).
The processing phase has enabled the creation of models with common checkpoints,
orthographic projections, sections and mapping of the anomalies with the indication of
levels and the tracking of the contour lines. An open and flexible planning marked these
activities, as considered essential to compare data acquired at different times, and frame
them into a single global coordinate system, independent from the tools and the methods
adopted. Different formats, in fact, must be able to be ‘read’ by various operating systems
and software as well as to allow a suitable data compression to reduce storage space and
to be consulted/managed online also by remote devices.
The comparison between the two models, geo-referenced into a single global system
based on 4 marked summits on the walls (which supposedly will remain unvaried over
the years) has permitted to ascertain the differences between relative movements of the
mosaic in the time elapsed between the two experiences. The anomalies have been
represented by the map of the relative deviations, considering only major differences of
±0.5 mm, as related to an effective geometric alteration and not the ‘disorders’ due to
the digital noise. The diagrams highlighted some punctual areas, with maximum values
of ±4–5 cm, showing a detachment of the tesserae and the consequent creation of new
gaps. Given the existence of a relatively rapid progression of the lifting phenomena, as
agreed with Villa’s management authorities, the operations will be repeated every four
months in the future. Following the replacement of the current shelter with the new one,
it would be also possible to check the eventual improvement of the mosaic conditions
that now requires preservation actions that cannot be anymore delayed. The analyses of
the geometry and of the discontinuities have allowed obtaining the framework of decay
actions in place but nothing says about the causes. It is an analysis on the surface
depicting the status of the mosaics but which provides no information on possible
reasons. The metric survey was then integrated with a microclimate monitoring and the
study of the physical and mechanical characteristics of the soil conducted by thermo‐
graphic surveys. The implemented diagnostic activities took into due consideration
previous analyses, which have allowed knowing the maximum environmental excur‐
sions, both daily and seasonally, while omitting an important factor: the ground temper‐
ature [17, 22, 23].
The current shelter of the triclinium, in fact, does not prevent the mosaics are directly
affected by sunlight and this determines high temperatures on the surface, as revealed
by these tests, of about 15°–20 °C higher than that of air. Measurements were performed
by placing the sensor - the first time at sunrise and again in the mid-afternoon - in
12 points at about 30–50 cm above the floor. Data were then processed by a gridding
and contouring software (for thermal maps and humidity distribution production) and
then superimposed to the orthographic projection. The comparison between ground
temperature values and the analysis of geometric deformations allowed verifying that
the most damage corresponds to thermal singularities. In detail, the campaign conducted
in the early hours of the morning showed that colder areas - due to increased walls
imbibition - are placed near the corners between the central and the two lateral apses.
Actually, on the left, masonry borders with a damp embankment (as evidenced by the
spontaneous and lush flora that covers it), while on the right side, there is a rain gutter
and a manhole, alas, both badly damaged. The distribution maps of the temperatures
measured at the hottest times, however, have led to the conclusion that because of the
absence of shielding in the entrance, the solar radiation, especially in the area close to
the peristyle, causes a strong surface’s heating, which extends along the longitudinal
axis until the central apse. This is, in fact, the most damaged portion of mosaic, even
because of the smaller resistance and elasticity of the mortar due to strong environmental
changes. The passive thermography confirmed the assumptions of the microclimate
monitoring and completed surveys in areas at risk. They have given clear indications
about the tesserae detachments, gaps and/or the presence of underground water [9]. Cold
areas in the zones where the geometric analysis had highlighted swellings, in fact, show
the presence of air between the backing and the tesserae’s layer. A dangerous situation
that, if not quickly and properly addressed, will accelerate the mosaics detachment
(Figs. 4 and 5).
Fig. 4. Metric survey: 3D rendering model and orthographic projection (survey 2015).
Fig. 5. Comparison between (a) elevations values and (b) altimetric deviations resulted from the
surveys of 2015 and 2016; (c) and (d) mapping of the ground temperature and (e) IR analysis.
5 Conclusions
The activities carried out on the triclinium of the Villa del Casale in Piazza Armerina
intend to provide a quick and low-cost methodology for the planned and preventive
checking of the mosaics preserved in situ. Certainly, only monitoring does not guarantee
the goods maintenance but can facilitate conservation status diagnosis thanks to the
structured recording of the knowledge acquired over time. It is the first step in a prepar‐
atory process towards prevention practices to implement before reaching a limit situa‐
tion. Diagnostic analyses conducted with light and relatively not too much expensive
instruments are suitable to be repeated at regular intervals, providing an updated over‐
view of the health status of the mosaics, useful to promptly direct any further investi‐
gations and therefore to make the right intervention. This will avoid incurring in super‐
ficial and hasty assessments and then in designing wrong solutions.
This procedure integrates the first critical analysis on the mosaics in its environment
with surveying - used in an innovative way - and diagnostics technologies. It develops
a codified path having specific timelines related to the intensities and dynamics of
examined objects and factors at play. The acquired data will then converge in a database
of the methods adopted and the results expected and achieved, normalized on shared
computerized language, to be easily updated and compared, so to verify the correctness
and effectiveness of project proposals, to direct and calibrate strategies and conservative
methods and any eventual change in design assumptions. In view of the spread of the
experimental data, it is to be highlighted the undoubted effectiveness of a web platform,
developed by diversifying the contents access from different users. The open-access to
diagnostic data acquired will be reserved only for specialists, so becoming a useful tool
for the site management1.
References
1. De Guichen, G.: La conservation préventive: un changement profond de mentalité. In: Study

series. Committee for Conservation (ICOM-CC), vol. 1, pp. 4–5. ICOM, Paris (2005)
2. Lambert, S.: Italy and the history of preventive conservation. CeROAr 6, 1–15 (2010)
3. Fiori, C., Riccobene, R., Tedeschi, C.: Tecniche di distacco nel restauro dei mosaici. Il Prato,
Padova (2008)
4. De Guichen, G., Nardi, R.: Mosaic conservation: fifty years of modern practice. In: Ben Abed,
A., Demas, M., Roby, T. (eds.) Lessons Learned: Reflecting on the Theory and Practice of
Mosaic Conservation, Proceedings of the 9th ICCM Conference, Hammamet, Tunisia,
November 29–December 3, pp. 9–14. Getty Conservation Institute, Los Angeles (2008)
5. Fiori, C., Vandini, M.: Teoria e tecniche per la conservazione del mosaico. Il Prato, Padova
(2002)
6. Basile, G.: Che cos’è il restauro, Ed. Riuniti, Roma (1989)
7. Della Torre, S. (ed.): La conservazione programmata del patrimonio storico architettonico.
Linee guida per il piano di manutenzione e il consuntivo. Guerini, Milano (2003)
8. Brandi, C.: Archeologia Siciliana. Bollettino dell’ICR 27–28, 93–100 (1956)
9. Michaelides D. (ed.): La conservazione. Uno strumento di conoscenza. In: 10a Conferenza
del comitato Internazionale per la Conservazione dei Mosaici (ICCM), Palermo, 20–26
October 2008. Regione Siciliana, Palermo (2014)
10. Gentili, G.V.: La villa romana di Piazza Armerina. Palazzo Erculio, Osimo (1999)
11. Settis, S.: Per l’interpretazione di Piazza Armerina. In: Mélanges de l’Ecole française de
Rome, Antiquité, vol. 87, no. 2, pp. 873–994 (1975)
12. Kähler, H.: Die Villa des Maxentius bei Piazza Armerina (Monumenta Artis Romanae xii).
G. Mann Verlag, Berlin (1973)
13. Vitale, M.R.: Un progetto “stratificato”. Pensieri e azioni per la tutela, il restauro e la
protezione della Villa romana del Casale di Piazza Armerina. In: Nigrelli, F.C., Vitale, M.R.
(eds.) Piazza Armerina. Dalla villa al parco. Studi e ricerche sulla villa romana del casale e
sul fiume Gela. Biblioteca del Cenide, Villa S. Giovanni, pp. 97–161 (2010)
14. Stanley-Price, N.: The Roman Villa at Piazza Armerina. In: de la Torre, M. (ed.) The
Conservation of Archaeological Sites in the Mediterranean Region, Los Angeles, pp. 65–84
(1997)
1
A. Versaci and A. Cardaci who shared methodological principles and finalities coordinated
this research. In detail, paragraphs 1 and 2 were developed by A. Versaci, 3 by L.R. Fauzia, 4
and 5 by A. Cardaci. Authors are very grateful to arch. Rosa Oliva, former director of the Villa
Romana del Casale Regional Museum for her trust in them and for her kind and precious
cooperation.
15. Oteri, M.A.: «a copertura l’azzurra volta del cielo». Il progetto di Piero Gazzola per la
conservazione a rudere della chiesa di S. Maria degli Alemanni a Messina (1940–41).
Quaderni del dipartimento patrimonio architettonico e urbanistico, 37–40, 223–230 (2010)
16. Scognamiglio, M.: Emergency intervention on flooded mosaic at Piazza Armerina. Newslett.
Int. Committee Conserv. Mosaics 9, 17–18 (1992)
17. Meli, G. (ed.): Progetto di recupero e conservazione della Villa Romana del Casale di Piazza
Armerina. I Quaderni di Palazzo Montalbo, vol. 12. Regione Siciliana, Palermo (2007)
18. Vivio, B.A.: Franco Minissi. Musei e restauri. La trasparenza come valore. Gangemi, Roma
(1991)
19. Ranellucci, S.: Conservazione e musealizzazione nei siti archeologici. Gangemi, Roma (2012)
20. Pensabene, P.: I mosaici della villa romana del Casale: distribuzione, programmi iconografici,
maestranze. In: Lentini, M.C. (ed.) Mosaici Mediterranei, pp. 87–116. Paruzzo, Caltanissetta
(2009)
21. Inglese, C.: Il rilievo integrato dei mosaici pavimentali nelle ville romane di Terme Vigliatore
e di piazza della Vittoria a Siracusa. Disegnare idee immagini 44, 80–91 (2012)
22. Bernardi, A.: Microclimate inside cultural heritage buildings. Il prato, Padova (2008)
23. Lucchi, E., Pracchi, V. (eds.): Efficienza energetica e patrimonio costruito. La sfida del
miglioramento delle prestazioni nell’edilizia storica. Maggioli Editore, Milano (2013)
Full Paper: Visualisation, VR and AR
Methods and Applications
A Mobile, AR Inside-Out Positional Tracking Algorithm,
(MARIOPOT), Suitable for Modern, Affordable
Cardboard-Style VR HMDs
Paul Zikas1,2 ✉ , Vasileios Bachlitzanakis1,2, Margarita Papaefthymiou1,2,

( )
and George Papagiannakis1,2

1
Foundation for Research and Technology Hellas, Heraklion, Greece
{mpapae02,papagian}@ics.forth.gr
2
University of Crete, Heraklion, Greece
{zikas,csd3058}@csd.uoc.gr
Abstract. Smartphone devices constitute a low-cost, mainstream and easy to

use h/w for VR rendering and main component for modern, mobile VR Head-
Mounted-Displays (HMDs). They support rotational tracking from on board
sensors to manage orientation changes, via their Inertial Measurement Units
(IMUs), but they lack positional tracking to reflect head translational movements,
a key feature that modern, desktop VR HMDs nowadays provide out-of-the-box.
Taking advantage of the RGB camera sensor that each modern mobile device is
equipped, we describe a novel combination of inside-out AR tracking algorithms
based on both marker and markerless tracking systems to provide the missing
positional tracking for mobile HMDs. We employed this system as an affordable,
low-cost VR visualization h/w and s/w method, for heritage professionals to
employ it for VR archeological sites and Cultural Heritage related monuments
interactive walk-throughs. We also compared our results with a recent holo‐
graphic AR headset (Meta AR-glasses) that supports gesture recognition and
interaction with the virtual objects via its RGB-D camera sensor and integrated
IMU.
Keywords: Mixed reality · Positional tracking · Monument visualization ·

Mobile VR · Marker and markerless tracking
1 Introduction
Most of Head Mount Displays (HMDs) support the basic three Degrees Of Freedom
(DOF) that includes roll, pitch and yaw. The calculation of those can be achieved by
onboard Inertial Measurement Unit (IMU) sensors which determine the orientation of
the camera in space. However, in order to track positional movements, we need an
additional external point of reference. This is usually implemented only in desktop VR
HMDs, with the use of a camera placed in the external environment (outside–in tracking)
or on-top or inside the HMD (inside-out tracking).
Positional tracking is the ability to determine the absolute position of the user’s HMD
within a three-dimensional space. By incorporating this feature into a VR headset we

DOI: 10.1007/978-3-319-48496-9_21
258 P. Zikas et al.
can represent these missing three more DOF in a total of 6. Originally Most HMDs could
track only head rotation, providing 3 DOF (e.g. Oculus DK1). The updated Oculus Rift
DK2 and current CV1 has an outside-in system for positional tracking with an external
camera placed on top of the monitor to track the position of the headset. Other commer‐
cial HMDs follow similar tracking approaches. Studies showed [14] that positional
tracking reduces the motion sickness these HMDs suffer as the orientation of the virtual
world is very similar to the real one. However, all current desktop-VR HMDs need to
be connected by cable to a computer. In contrast, smartphones provide a mainstream
and easy to use mobile VR platform but they lack positional tracking. Visual Markers
seems ideal for such low cost, mobile systems [16] that need a distinct visual sing to
determine a static position from camera. By placing a marker in front of our desk we
are simply define a static point in the real world in order to collect information about
the relative position of the marker and our camera. Although markers are easy to use it
is not very convenient to attach a marker everywhere we need positional tracking
features. To overcome this limitation, we can use a markerless, SLAM-based AR
tracking system. The main principles remain the same as we have to scan the environ‐
ment to track visual features and extract a 3D map of the surroundings. Later we will
use this mapping as a 3D marker to calculate the position of the camera like before.
For the above reasons we introduce two different ways of positional tracking with
both marker and markerless implementations. For the needs of this project we developed
a virtual reality navigation in the Palace of Knossos. User can navigate his way through
the archeological site by rotating his head and leaning back and forwards or left and
right as positional tracking is supported. To the best of our knowledge this is the first
time in bibliography that AR inside-out positional tracking enhance conventional IMU
sensor tracking in mobile VR.
In the following sections we present MARIOPOT for both marker and markerless
implementations. In more detail, in Sect. 2 we describe the previous work in the area
indicating AR and VR applications which appeared useful. Afterwards (Sect. 3) we fully
describe MARIOPOT for both approaches (marker and markerless) by presenting the
methodology, the needed calculations and the used matrixes. The next two paragraphs
describe both the marker and markerless implementation of our application by pointing
out the differences in each approach. In Sect. 6 we compare our results with Meta AR-
glasses using marker based positional tracking with and gesture based interaction to
handle and examine the archeological site of Knossos. Finally, we present the conclu‐
sions of our work and the future work-research that can be done.
2 Previous Work
When reconstructing a digital model of an archeological site it needs to be as accurate

as possible [1] to improve the user experience (UX). For this reason, the model of
Knossos we used is a realistic representation of the Palace as it stands today [11]. In
recent years VR approaches for educational purposes have introduced a didactic poten‐
tial in the area of cultural heritage. As [3] states, there are various VR systems available
for cultural heritage purposes to enhance the interest of the general public. The benefit
A Mobile, AR Inside-Out Positional Tracking Algorithm, (MARIOPOT) 259
of an immersive and interactive simulation imparts knowledge and further motivation

for great interest and research in the area of cultural heritage.
Both marker and markerless tracking intensify the basic features of augmented
reality applications. [2] presents an AR application with markerless tracking about a
quiz based on a museum visit. Another notable approach was done in [10] where a
markerless system with automatically recover from features loss, runs an AR application
for the maintenance of a car’s engine. A robust authoring of life-sized AR characters
for mobile devices was presented in [12], where a fast pipeline based on Metaio was
used to populate AR scenes. The work featured an easy to use method for rendering AR
characters with a novel pipeline in under one-minute process. The tracking of the area
was done beforehand with Metaio to export a 3D mapping of the scene that will be later
used for the character’s authoring. The augmented characters were able to perform
various animations, gestures and speech with the use of SmartBody library. This SLAM
tracking method allows the augmentation of any indoors or outdoors scene capable of
generate a decent number of features that will export the 3D map.
The importance of an easy to use AR mobile tracking system is emphasized in
[7]. Daniel Wagner et al. present an accurate 6DOF pose tracking system that tracks
conventional markers to render augmented content. Since most mobile devices have
a build in camera they provide a complete tracking platform able to manage marker
based tracking with the least effort. There is also an interesting apposition of the
performance each mobile device had after a series of tests, providing a more tech‐
nical view of the research. Moreover, [13] proposed a robust tracking system to
determine the position of a mobile device in space using the build-in GPS sensor. A
careful study on the energy consumption of different GPS based techniques is
presented in parallel with their accuracy and efficiency. Mobile devices have limited
power supplies requiring such systems to tackle this issue and manage the energy
consumption to the point they reduce their impact on it.
In this work we have used the open source OpenGL Geometric Application (glGA)
framework [8] implemented for IOS mobile devices. More specifically, glGA frame‐
work is a shader based computer graphics framework supported by desktops and mobile
devices. The architectural structure of glGA provides the basic functionalities of
OpenGL applications like loading and rendering both static and rigged 3D meshes,
compiling and loading shaders, load textures and play animations.
3 Our Mobile, AR Inside-Out Positional Tracking (MARIOPOT)

Algorithm
For the complete 6 DOF movement we need two main components: (1) The rotation of
our device to determine the orientation of HMD in space and (2) its position to compute
the translation in the digital scene. In our application we integrate the basic functional‐
ities of cardboard SDK to get the rotation values as reflected from the orientation sensors.
These values appear as a 3 × 3 matrix that from which we constructed the final view
matrix of our application. Except from the rotations we also need the camera’s transla‐
tion. This is the reason we used a marker to provide the translation values. The position
260 P. Zikas et al.
of the camera is a vector with three elements that defines the translation of the camera
in all three axes.
From the extraction of these two components we have to construct the view matrix
of the application. To make the transformations work properly we should not apply the
matrix directly to our object as this will cause malfunctions in the visualization process,
especially when the marker is no longer in sight. We have to construct a custom LookAt
matrix from the extracted values as we need to rotate the world from the point of camera
and not from the position of Knossos. By doing this we eliminate an issue that positional
tracking brings about when the camera is changing position. We have to take in consid‐
eration the position of the camera in every frame as we always want to rotate the scene
according to this specific point.
Below there is a brief explanation of the needed calculations to export the final view
matrix.
Camera’s Position. The position of the mobile device we are using can be represented
as a three-dimensional vector. In both cases (marker and markerless) we extract the
positional vector from the computed model matrices OpenCV and Metaio provide,
saving only the last column of the matrix which holds the camera’s position.
For the Marker implementation we used the detection mechanism [6] proposed to
get the transformation values. After we calculate the marker’s position we use the posi‐
tional values to generate the application’s Model Matrix as seen below.
The Markerless approach differentiate on the calculation of positional values made

by Metaio. The code below shows the construction of the transformation matrix in each
frame.
Camera’s Rotation. The next step is to calculate the orientation of our mobile device.
For both implementations we used CardboardSDK to extract the desired rotational
values. The SDK provides build-in functions to get the head matrix of cardboard that
represents the orientation of our device.
From the rotational matrix we have to compute three vectors that will constitute the
orientation of our camera in all three axes. In order to create a functional camera matrix,
we also need to determine the eye position which describes the position of our camera
in virtual space. The eye position is defined from the rotational values and the position
of our camera as follows.
eyex = −dot(xaxis, position)

eyey = −dot(yaxis, position)
eyez = −dot(zaxis, position)
Final Matrix. The final LookAt matrix is computed by combining the above vectors of
rotational and positional values. Our custom LookAt matrix will be used to compute the
final view matrix of our application. It is important to notice that we have to take care
of the matrix order while working with matrices from different systems or SDKs as the
default order for OpenGL is column-major whereas Cardboard represents matrices in
row-major order (Fig. 1).
However, there is a major issue with inside-out tracking when combining rotational
tracking from cardboard and positional tracking using marker of markerless implemen‐
tation. When we rotate the camera of our device we have two data inputs: (1) Cardboard
SDK calculates the change of mobile’s orientation. (2) From the point of marker it seems
the model has changed position. This not accurate as we have a conflict between these
262 P. Zikas et al.
Fig. 1. Positional tracking in both marker and markerless implementation.
two data streams. Only if we use the marker for both rotational and positional tracking
the visualization seems correct. To make things more clear, if we keep our device static
and rotate it to the left while keeping contact with the marker, our virtual world will
simultaneously rotate to the right (from Cardboard) and translate to the right (from
marker) as from the camera’s feedback the marker seems to move also to right. This is
incorrect as we want our world to remain static but the camera will catch the marker
moving to right and change the position of our virtual camera too. To reduce this effect
we have to separate the two movements (rotation and translation) for both marker and
markerless implementation. We provide a demonstration of this implementation in the
next section (Fig. 2).
Fig. 2. The architectural diagram of our application
4 AR Positional Tracking with Markers
In order to determine the position of the camera we need to define a reference point in
space. Marker based tracking is an efficient, low cost and easy to use method to achieve
this result. To detect possible markers, we need to enable the camera of our mobile
device and collect the data from live image streaming. Since every mobile device has
an onboard camera there is no need for an additional sensor or active tracker to determine
the position of the device. We used OpenCV image recognition patterns [6] to compute
the position of our maker [9]. The detection of markers requires a pipeline of image
processing that begins with the definition of existing candidates and finishes with the
determination of the actual markers. The vector extracted from this method containing
the positions in three-dimensional space will set the camera’s translation in our virtual
scene (Fig. 3).
Fig. 3. Setting the marker and our HMD for marker based positional tracking
Marker based tracking has the limitation of a reference point that always need to be
in camera’s sight. If the marker is no longer in contact with camera the mobile device
will stop calculating its position in the virtual scene as there will be no point of reference
in the real world. The small field of view most mobile phones have will cause even more
marker losses after an amount of rotation or translation. In our case, when the marker is
out of reach of camera, the rotational tracking is applied to the last known position of
the mobile device. This is achieved by saving the position we lost track with the marker
and continue calculate the orientation values from the cardboard SDK. This implemen‐
tation provides a smooth rotation even if the marker is not in sight and in the situation
of the marker loss, the application continues to be functional with rotational tracking.
When the camera tracks again the marker, we start computing again the transformation
values and the positional tracking is enabled from the new position.
As mentioned before, we have to separate the positional and the rotational tracking
to perform rotations without unwanted positional changes by the moving marker. To
manage this issue we had to deactivate the marker when our mobile device overcomes
a fixed angular threshold. By doing this the virtual scene will stop moving in a static
rotation indicated incorrectly from the marker as the marker will no longer affect the
position of the camera. We will activate the marker again when the mobile device reaches
again lower angles and the computation of positional tracking will start again.
5 Markerless Positional Tracking
Unlike conventional tracking with markers, markerless tracking is definitely more flex‐
ible and reliable. To implement markerless AR camera tracking we employed Metaio
SDK and the Toolbox. Both applications make markerless tracking an easy task by
simplifying the process of capturing features and generating the final view matrix of the
visualization. The procedure of markerless tracking is more complicated than placing a
marker in front of your desk but it can track large areas (e.g. small rooms) providing
user with more space to move around. With the markerless approach we have the oppor‐
tunity to leave our desk to walk freely in a room and still receiving positional feedback
from the 3D mapping we tracked (Fig. 4).
264 P. Zikas et al.
Fig. 4. The process from capturing features with Toolbox to markerless positional tracking.
To implement markerless tracking we need a 3D map of the area we are about to run
the application. A 3D map is actually a file that contains the positions of district features
in the environment and works the same way as a typical marker. We create this file by
using Toolbox application. To make the file we run the application and scan the envi‐
ronment to capture visual features, the more features the better. After saving the 3D map
we have to transfer it to the same folder our VR applications runs in order to detect the
physical environment as a 3D marker.
Markerless implementation for positional tracking has the same methodology as
marker based ones. We have to calculate again the rotational matrix form cardboard
SDK and the translation matrix from Metaio SDK. The positional vector extracted from
Metaio it is still a three-dimensional vector but the origin of tracking in 3D space is
represented by a 3D marker (e.g. a box on top of our desk) and not by a 2D sign as
marker based approach does. However, with markerless tracking we have a greater area
of possible movements thus we have to scale the translation matrix accordingly to main‐
tain the proportions of real and the virtual world translations. When the camera loses
visibility of the features we maintain the last known position of the mobile phone and
from there user can perform head rotations without having the ability to move to another
position. After features become visible again user can start again moving around the
virtual scene.
6 Comparison with Meta AR Glasses
Since we had our first results from the positional tracking we implemented this method
with a different HMD that utilizes an RGB-D camera sensor. The main motivation
behind this comparison is due to the fact that many forthcoming smartphones will feature
such an RGB-D camera as part of their standard configuration h/w and we wanted to be
the first to study their use and draw a comparison.
We reproduced our methodology in Meta AR-glasses, a holographic, see-through
headset with gesture recognition. Wearing Meta AR-glasses gives user the ability to
interact with holographic objects by using basic gestures [5]. Holographic headsets
constitute an ideal HMD to experiment with novel AR applications [4]. Meta AR-glasses
support rotational tracking from onboard sensors but they lack of positional tracking
due to the absence of an external camera. However, we can utilize the embedded marker
recognition mechanism to introduce positional tracking functionality (Fig. 5).
Fig. 5. Interaction using markers and hand gestures.
Meta use markers to attach holograms onto them but we used them in a different
way. As mentioned earlier, we need to have a point of reference in the real world to
manage the translations in the virtual one. To implement this idea we have to attach our
virtual scene to the marker. By doing this we will enable positional tracking for Meta
glasses. We need to compare applications with similar content to have more accurate
results so for the Meta application we used again the model of Knossos. As indicated in
[15] the two technologies have some basic differences in immersion, interaction and in
the scale of the digital environment but the tracking mechanism remains the same. The
positional tracking was of the same quality as it was implemented with the same prin‐
ciples. In general, by introducing positional tracking is Meta AR-glasses we were able
to zoom in and out the palace of Knossos and see the model in more detail than previously
when the only way to handle the model was with hand gestures.
While mobile VR applications have the advantage of generating an immersive expe‐
rience, Meta headset with gesture recognition enhances the interaction between user and
the hologram he sees. Interaction in game based applications has great importance as
user is no longer an observer, he is able to handle holographic objects with native hand
movements. In cultural heritage we can use this feature of Meta AR-glasses to handle
and examine known monuments or important buildings from a different perspective.
In order to compare the user experience of Meta glasses and our mobile tracking
system, we conducted an experiment to evaluate the use of both technologies. The
participants were 7 in total (6 male, 1 female). For the mobile orientation demo partic‐
ipants had to navigate through the palace of Knossos by making head movements. In
this way we were able to test if the navigation system was simple and easy to use. Most
of the participants performed well as they managed to find their way in the ruins within
less than a minute. After a while they were more confident and start to explore the
archeological site in detail by zooming at point of great interest (e.g. the Prince with the
lilies fresco). However, half of the participants complained about dizziness which was
something expected as it takes time to get used to virtual reality. Another think that needs
to be mentioned is that most of the participants had problems with the interruption of
positional tracking caused by the loss of visual contact with the marker. As was shown
by the process this was the most common issue we faced at the experiment.
For Meta AR glasses the participants had to handle a building of the archeological
site with the marker provided. In addition to the previous experiment we enhanced the
266 P. Zikas et al.
user experience with gesture based interaction to find out if there will be any significant
change to the final remarks. We asked the participants to zoom in specific points of the
building as well to rotate, translate and scale it with their hands. Most of the participants
had a serious problem with gesture handling as they couldn’t manage to grab the building
correctly. Besides this, the marker based handling of the building was a task that
completed successfully from all the participants. About this, there were few complains
referring to the limited field of view Meta AR glasses provide as the building was
cropped from being too close to the headset leading to the failure of the immersion. As
a final remark, the participants were satisfied from the HMD’s orientation capabilities
that were enhanced with the use of positional marker based tracking.
To conclude, both HMDs perform well in the positional tracking and orientation in
general. Our mobile implementation had better results considering the immersion of the
archeological site as the participants were in a fully virtual environment unlike Meta
glasses. Since we used conventional markers in both methods to provide positional
tracking they had very similar results in accuracy and performance. An extra feature we
used with Meta headset was the gesture handling they support but we faced poor results
as it was difficult for most of our participants to successfully perform the correct gestures
for the device to recognize (Fig. 6).
Fig. 6. Comparison table of MARIOPOT and Meta AR glasses.
In this work we presented a low cost and easy to use implementation of positional
tracking for mobile device-based, cardboard-style VR HMDs that are ideal to be used
as affordable visualization for cultural heritage professionals and sites. It is the first time
in bibliography that such a hybrid orientation system is presented, combining sensor and
camera tracking. With the contribution of positional on-top of rotational tracking in
mobile VR, we were able to zoom in and out not only to change our orientation but also
our translational position in a virtual world and thus appreciate the presented 3D monu‐
ments in better detail, with less motion sickness effects. The presented algorithm can
improve the visualization of digitized archeological finds (e.g. pottery, frescos, and
coins) instead of using conventional software methods. Since the digitization process in
Cultural Heritage is used frequently, we have to develop the appropriate tools to make
better use of those findings. A VR demonstration of monuments and historical buildings
is without a doubt the best way of visualization as the immersion generated from this
technology can really make the difference. Mobile AR, inside-out Positional tracking
can extend the basic limits of VR and especially when implemented with markerless
tracking in room sized areas as user will not only be an observer of the monument but
he will actually walk and explore the site on his own. Such technology can be used in
museum galleries and expeditions to attract the interest of general public and to inform
users about the importance and the benefits of digital preservation.
In the future we aim to replace the Metaio SLAM-based markerless tracking with
OpenCV. We are also planning to enhance the visualization of Knossos with gamifica‐
tion elements to constitute a fully interactive experience of the archeological site. Serious
game industry expands rapidly over the last years introducing new ways of learning
through gaming. Cultural heritage applications can definitely benefit from a more inter‐
active environment that will augment their efficiency.
Acknowledgements. The research leading to these results has received partial funding from the
European Union People Programme (FP7- PEOPLE- 2013-ITN) under grant agreement No.
608013 and from the European Union Horizon2020 Programme (H2020-SC6-CULT-COOP-9 –
CSA) under grant agreement No. 727107. Special thanks also to ARdictive (www.ardictive.com)
and its CEO & founder Nils Huebner for the MetaAR glasses.
References
1. Foni, A., Papagiannakis, G., Magnenat-Thalmann, N.: A taxonomy of visualization strategies

for cultural heritage applications. ACM J. Comput. Cult. Heritage 3(1), 1–21 (2010)
2. Yudiantika, A.R., Sulistyo, S., Hantono, B.S.: The development of mobile augmented reality
quiz visualization methods based on markerless tracking for museum learning application.
Disampaikan pada seminar “The International Forum on Strategic Technology (IFOST)”,
June 2015
3. Gaitatzes, A., Christopoulos, D., Roussou, M.: Reviving the past: cultural heritage meets
virtual reality. In: Proceedings of the 2001 Conference on Virtual Reality, Archeology, and
Cultural Heritage (VAST 2001), pp. 103–110. ACM, New York (2001)
4. Billinghurst, M., Clark, A., Lee, G.: A survey of augmented reality. Found. Trends Hum.
Comput. Interact. 8(1), 1–202 (2015)
5. Billinghurst, M.: Gesture based AR interaction research at the HIT Lab NZ. Presentation
(2015)
6. Baggio, D.L., Emami, S., Escrivá, D.M., Ievgen, K., Mahmood, N., Saragih, J., Shilkrot, R.:
Mastering OpenCV with Practical Computer Vision Projects
7. Wagner, D., Schmalstieg, D.: ARToolKitPlus for pose tracking on mobile devices. In:
Grabner, M., Grabner, H. (eds.) Computer Vision Winter Workshop 2007, St. Lambrecht,
Austria, 6–8 February 2007. Graz Technical University (2007)
8. Papagiannakis, G., Papanikolaou, P., Greasidou, E., Trahanias, P.: glGA: an OpenGL
geometric application framework for a modern, shader-based computer graphics curriculum.
In: Eurographics 2014, pp. 1–8 (2014)
9. Kato, H., Billinghurst, M.: Marker tracking and HMD calibration for a video-based augmented
reality conferencing system. In: Proceedings of the IEEE International Workshop on
Augmented Reality, pp. 125–133 (1999)
268 P. Zikas et al.
10. Platonov, J., Heibel, H., Meier, P., Grollmann, B.: A mobile markerless AR system for
maintenance and repair. In: Proceedings of the 5th IEEE and ACM International Symposium
on Mixed and Augmented Reality (ISMAR 2006), pp. 105–108. IEEE Computer Society,
Washington, DC (2006)
11. Kateros, S., Georgiou, S., Papaefthymiou, M., Papagiannakis, G., Tsioumas, M.: A
comparison of gamified, immersive VR curation methods for enhanced presence and human-
computer interaction in digital humanities. Int. J. Heritage Digit. Era 4(2), 221–233 (2015)
12. Papaefthymiou, M., Feng, A., Shapiro, A., Papagiannakis, G.: A fast and robust pipeline for
populating mobile AR scenes with gamified virtual characters. In: SIGGRAPH Asia 2015
Mobile Graphics and Interactive Applications (SA 2015). ACM, New York (2015). Article
22, 8 pages
13. Kjærgaard, M.B., Langdal, J., Godsk, T., Toftkjær, T.: EnTracked: energy-efficient robust
position tracking for mobile devices. In: Proceedings of the 7th International Conference on
Mobile Systems, Applications, and Services (MobiSys 2009), pp. 221–234. ACM, New York
(2009)
14. Desai, P. R., Desai, P. N., Ajmera, K. D., Mehta, K.: Int. J. Eng. Trends Technol. (IJETT)
13(4) (2014)
15. Zikas, P., Bachlitzanakis, V., Papaefthymiou, M., Kateros, S., Georgiou, S., Lydatakis, N.,
Papagiannakis, G.: Mixed reality serious games and gamification for smart education. In: 10th
European Conference on Games Based Learning (ECGBL), October 2016
16. Zhang, X., Fronz, S., Navab, N.: Visual marker detection and decoding in AR systems: a
comparative study. In: Proceedings of the 1st International Symposium on Mixed and
Augmented Reality (ISMAR 2002), p. 97. IEEE Computer Society, Washington, DC (2002)
Virtual and Augmented Reality Tools to Improve
the Exploitation of Underwater Archaeological Sites
by Diver and Non-diver Tourists
Fabio Bruno1 ✉ , Antonio Lagudi1, Loris Barbieri1, Maurizio Muzzupappa1,

( )
Gerardo Ritacco1, Alessandro Cozza2, Marco Cozza2, Raffaele Peluso2,

Marco Lupia3, and Gianni Cario3
1
Department of Mechanical, Energy and Management Engineering (DIMEG),
University of Calabria, Rende, CS, Italy
{f.bruno,antonio.lagudi,loris.barbieri,
muzzupappa,gerardo.ritacco}@unical.it
2
3D Research S.r.l., Rende, CS, Italy
{a.cozza,m.cozza,r.peluso}@3dresearch.it
3
Applicon S.r.l., Rende, CS, Italy
{m.lupia,g.cario}@applicon.it
Abstract. The underwater cultural heritage is an immeasurable archaeological

and historical resource with huge, but yet largely unexploited, potentials for the
maritime and coastal tourism.
In this regard, in the last years, national and international government author‐
ities are supporting and strengthening research activities and development strat‐
egies, plans and policies to realize a more sustainable, responsible and accessible
exploitation of the underwater cultural heritage.
To this end, the paper presents the architecture of a new system that, taking
advantage of the modern virtual and augmented reality technologies, allows diver
and non-diver tourists to make a more engaging and educational experience of
the underwater archaeological sites.
This system has been developed and tested in the VISAS project (www.visas-
project.eu) that aims to the enhancement of the cultural and tourist offer related
to the underwater archaeology through innovation of modes of experience, both
on site and remote, of the underwater environments of archaeological interest.
Keywords: Underwater cultural heritage · VR systems · AR systems ·

Underwater 3D reconstruction
1 Introduction
The underwater cultural heritage (UCH) is an extensive and varied asset with wide
potentials for the development, but in return it needs protection, care and investment.
These potentials could be exploited in a fruitful and efficient way especially in the
tourism sector, which enables high profit margins for an unlimited range of time. But
most importantly, in addition to the economic return, the exploitation of the UCH

DOI: 10.1007/978-3-319-48496-9_22
270 F. Bruno et al.
provides also an immediate value creation in terms of entertainment and learning expe‐
rience, for the cultural enrichment of the individual, and of reputation and notoriety of
a location, for the local identity that makes a place unique and can direct tourism streams.
The tourism is clearly an important economic activity, especially in many maritime
and coastal regions. In fact, every EUR invested in heritage increases economic activity
around it by up to 12, on the contrary, heritage destruction decrease the value of its
location [1].
Considerable improvements are going on in the UCH field especially thanks to an
articulated implementation of the Unesco conventions 1970 (prevention of illicit traffic),
1972 (World Cultural and National Heritage), and 2001 (Underwater Cultural Heritage).
In particular, the 2001 Unesco Convention on the Protection of the Underwater Cultural
Heritage [2] has represented important achievements in the field of submerged archaeo‐
logical site safeguarding that have led to the rise of a plethora of innovative methods
and tools that cover a wide range of applications [3–8] for this burgeoning sector.
In addition to the work carried out by National Commissions for Unesco other
national and international organizations and government bodies are already moving in
the same direction to make the underwater archaeological sites’ exploitation more
sustainable and accessible to large-scale tourism. It is worth to mention the European
Commission that has set-up the EU’s Blue Growth strategy [9] for the cooperation
among research institutes, museums, tourism companies and other stakeholders in order
to develop innovative and sustainable solutions and products, with a maximum use of
information technology, to boost the tourism in coastal destinations.
In accordance to these recommendations, the paper presents a new system that,
taking advantage of the modern virtual and augmented reality technologies, allows diver
and non-diver tourists to make a more engaging and educational experience of the
underwater archaeological sites. In fact, this system, on the one hand, promotes the
diving tourism by improving the divers’ experience in the underwater site, and, on the
other hand, promotes the induced tourist activity through the development of an inno‐
vative virtual tour of the site.
In particular, a virtual reality (VR) tool allows non-diver tourists to live a virtual
experience inside the reconstructed 3D model of the underwater archaeological site. The
VR tool provides also an interactive navigation within the virtual environment to plan
the guided dives. Instead, an augmented reality (AR) tool is intended for diver tourists
that visit the underwater site, allowing them to have a virtual guide that provides specific
information about the artifacts and the area they are visiting. The AR tool is based on a
tablet properly equipped with a waterproof case and an integrated system for acoustic
localization and inertial navigation.
This system has been developed and tested in the VISAS project (Virtual and
augmented exploitation of Submerged Archaeological Sites - http://www.visas-
project.eu) that is a collaborative research project funded by MIUR (Italian Ministry of
Education, University and Research); it has started on 1st April 2014 and will end on
30th September 2016.
The VISAS project aims to the development of tools for improving the visitor’s
experience of underwater archaeological sites by making it more interesting, charming,
and effective [10]. The VISAS project architecture is depicted in Fig. 1. The process
Virtual and Augmented Reality Tools 271
starts with the acquisition of optical and acoustic data. The first can be carried out by
scuba divers or by Remotely Operated underwater Vehicles (ROVs), the second by
multibeam sonar system mounted on the bottom of a survey vessel. Downstream of the
process, there is the opto-acoustic 3D reconstruction from which the multi-resolution
textured 3D models of underwater archaeological sites are generated. These models are
stored into a database and made available to the diver and non-diver tourists by means
of the VR and AR diving tools.
Fig. 1. VISAS project architecture

272 F. Bruno et al.
2 UCH Exploitation by Means of VR and AR Technologies
The virtual reality and augmented reality technologies have demonstrated to ensure a
good effectiveness in increasing the value of cultural heritage [11–15], but the possible
applications for underwater archaeology have not been sufficiently investigated. The
potentialities of these 3D graphics techniques in the cultural heritage exploitation are
related to the possibility to de-materialize the heritage itself, so that it can be enjoyed
without any constraint given by distance or time, giving also the possibility to live an
impressive learning experience with a high emotional impact. These potentialities may
be also more important for the exploitation of the underwater cultural and naturalistic
heritage, because its accessibility is limited to diver tourists.
For a few years now, various researches are investigating and proposing different
frameworks for the reconstruction, collection and visualization of the UCH but these
results limit the exploitation to the single underwater archaeological remain [16] or are
more oriented to the digitization for scientific purposes [17–21] instead for the tourist
exploitation.
This lack is more evident in the case of AR technology, in fact, at the moment there
is only a research that proposes the adoption of an underwater tablet but its usage is
limited to the geolocation and diver monitoring [22].
As a result, the VR and AR applications for the exploitation of the UCH are still few,
with many shortcomings and huge development potentials.
3 Opto-acoustic 3D Reconstruction
The 3D reconstruction of the underwater archaeological site is worth a particular atten‐

tion because of the long and articulated process. In fact, the creation of a multi-resolution
textured 3D model of the underwater archaeological site requires various steps with an
integrated adoption of both optical and acoustic techniques [23–26]. In particular, the
VISAS project proposes a method to merge the high resolution data acquired through
photogrammetric techniques with acoustic microbathymetric maps to obtain a complete
representation of the underwater scene and to geo-localize the optical 3D model using
the acoustic bathymetric map as a reference.
After a first inspection of the site and a calibration of the optical and acoustic equip‐
ment, the photogrammetric acquisition process is performed according to standard aerial
photography layouts that consist of overlapping straight lines and also cross lines with
oblique poses to minimize the occluded areas.
The acoustic acquisition process is performed thanks to a high frequency multibeam
equipment that allows to obtain an acoustic bathymetry of the seabed.
Custom opto-acoustic markers, placed on the seabed and whose number depends by
the extension of the site, are used to accurately compute the registration between the
optical and acoustic point clouds. While, a set of triangular target are adopted to scale
the optical 3D model.
The last steps of the process consist of meshing and texturing the opto-acoustic point
cloud of the underwater archaeological site. The meshing step is carried out using a
dedicated software, which has the ability to create a mesh by using an efficient multi-
resolution algorithm and to perform further refinements of the model by using the point
cloud as reference, so that the model reconstruction is performed in a coarse-to-fine
fashion.
A technique based on the projection and blending of 2D images on the 3D surface has
been adopted in order to place textures on the 3D model. In particular, since the camera
poses are known downstream of the optical 3D reconstruction process, the high resolution
images are mapped on the portion of the 3D surface of the model representing the archaeo‐
logical remains. The low-resolution polygonal mesh of the seabed is in turn obtained from
the acoustic bathymetry and textured with a tile-based texture mapping approach that just
requires to set some sample images of texture tiles instead of a large texture.
4 VR Diving Tool
The VR tool is a virtual diving exhibit that allows non-diver tourists to live a virtual
experience and learn both general information and historical-cultural contents related
to the specific archaeological site. In particular, thanks to the VR tool, tourists are able
to explore the 3D reconstruction of the underwater site and to receive historical and
archaeological information about the submerged exhibits and structures of the site, but
flora and fauna are also described, with a particular attention on their interaction with
the submerged artifacts.
The VR tool can be also used by diver tourists because of its capability to make a
detailed planning of the operations and of the itinerary to carry out in the underwater
archaeological site. The tool indeed represents a reliable instrument to plan and simulate
the tourist itinerary that is performed at a later time in the real submerged environment.
The VR diving tool has been implemented by means of the cross-platform game
engine Unity in order to take full advantage of its capability to program software that
can be used and communicate on the web. In fact, a web service software represents the
key element of the software architecture because it provides the linking between the
database and a scene development environment. The development environment is
composed of two modules, i.e., an editor and an interaction module. The scene editor
module allows to compose the virtual scene by integrating 3D objects and multimedia
information stored in the database. The main elements of a virtual scene are: the 3D
texturized model of the underwater archaeological site; 3D models of the flora and fauna;
the Points of Interest (POI) that contain educational textual, visual and audio information
related to a specific underwater area or a specific kind of archaeological artifact. Further‐
more, depending on the type of the element, other metadata are integrated into the scene,
such as the color and luminosity for the lights. Once the scene is created the interaction
module is adopted to implement the logics of the virtual scenario defining the physics
of the elements that belongs to the scene. Furthermore, it loads from the database the
graphical assets of the submerged, terrestrial and aerial environments, such as, refrac‐
tions, fog, caustics of the particulate, etc.
The interaction module is used also to perform the exploration within the virtual
scenario according to the user input get by means of the controller.
274 F. Bruno et al.
From the user’s point of view, the VR diving tool presents two different possible
embodiments each one characterized by the type of devices, the provided levels of
immersion, interaction and presence.
4.1 Semi-immersive VR Environment

The VR tool can involve tourists in a semi-immersive visualization by means of a full
HD monitor based on passive 3D technology (Fig. 2). The passive technology has been
preferred to the active one because active 3D glasses are expensive and need batteries
to work. Furthermore, passive 3D glasses are inexpensive, lighter and more comfortable.
Fig. 2. Tourist engaging with the VR tool in the semi-immersive environment (a). POI’s
information provided by the VR diving tool (b)
Tourists interact with the system by means of a multi-touch tablet featuring a user
interface (UI) that provides all the input functionalities needed to explore the 3D envi‐
ronment and get access to the multimedia data. In particular, as shown in Fig. 2a, the UI
provides to the tourist two large command buttons, respectively, to go back and forth
and to rotate the camera’s point of view. While, on the top left side of the UI, a slider
controls the depth of the camera view from the water surface. A set of option buttons,
placed in the center of the UI, are dedicated to the dive planning. In particular, the tourist
can load or close a path and play, pause, or resume an automatic flight of the planned
dive session which is later performed in the real underwater archaeological site. On the
top part of the UI a detail disclosure button opens a modal view, depicted in Fig. 2b,
containing additional textual, graphical and audio information related to a specific POI.
4.2 Immersive VR Environment
The immersive diving environment is provided by the VR tool to the tourists by means
of HMD (head mounted display) technology. The HMD isolates the user from the
distractions of the actual physical environment and encompasses the entire field of view,
including the peripheral space. The navigation in the virtual environment is performed
by the tourist by moving his/her head and interacting with a single wireless handheld
controller (Fig. 3). In particular, a HTC Vive [27] has been used to implement the
immersive environment.
Fig. 3. Tourist engaging with the VR tool in the immersive environment
When the tourist wears the HMD he/she experiences the immersive virtual environ‐
ment from the scuba diving viewpoint simulating a real diving session. Then the scenario
that appears to the tourist, at the beginning of the virtual experience, is above the water
surface in the point of immersion. Once he/she dives in the submerged virtual environ‐
ment, he/she is guided by a directional 3D arrow (Fig. 3a) to the archaeological under‐
water site. When the tourist arrives to the site the 3D arrow disappears and lets him/her
free to interact with the 3D POIs to discover historical and archaeological information
(Fig. 3b).
Compared to the semi-immersive environment, which relies on monitors for the
visualization, in the immersive environment tourists receive audio contents instead of
visual information when interacting with 3D objects and POIs.
5 AR Diving Tool
The AR tool consists in an underwater tablet equipped with an underwater positioning

and orientation system that guides the diver tourists during the diving session while
providing information about the archaeological artifacts spotted during the visit.
The underwater positioning and orientation system is performed by means of an
acoustic modem, integrated to the underwater tablet, which estimate the position of the
receiver by computing the distance from at least three fixed transmitters (beacons) placed
on the seabed (LBL - Long BaseLine technique) (Fig. 4). In order to improve the accuracy
and increase the robustness in case of loss of signal from one or more beacons, the tablet is
also equipped with an inertial platform and a depth sensor. The data coming from the
various sensors are processed through data fusion and error estimation algorithms.
The tracking system on the tablet sends a query to the beacons and computes the
distance from each of them. These data are used by the data fusion algorithm to correct
the estimate on the position obtained through the inertial system and the pressure sensor.
276 F. Bruno et al.
Fig. 4. Augmented diving system architecture
The navigation software receives this information and shows the location of the tablet
on a 3D map of the underwater archaeological site.
Each beacon contains an electronic board to drive the piezoelectric transducer. It
operates at 25–30 kHz frequency band. The beacon is equipped with a battery pack,
which enables up to five hours of use, and with anchor rings used to fix it on the buoy
cable. The beacons are positioned in known geographical points using a mooring post
and a surface buoy that facilitates their recovery.
The tablet is composed of two major parts: a fully functional underwater touchscreen
housing and a waterproof case for the electronics of the tracking system. The two devices
are connected using a Bluetooth Low Energy (BLE) interface.
The tablet updates the diver’s position on the 3D map at a frequency of 50 Hz. The
current position is estimated through an Extended Kalman Filter that uses the distances
from each beacon, the depth provided by a pressure sensor, the accelerations along the
three axes (x, y, z), and the orientations provided by an inertial platform.
In order to compute the distance between the tablet and a beacon, the simplest way
is to compute the time-of-fly of the acoustic wave. To this end, the acoustic modem on
the tablet sends a two-way range (TWR) command to the remote beacon and starts an
internal timer. When the tablet receives the response message sent by the beacon, the
timer is stopped. The position system sends cyclically a TWR to all beacons. The tablet
position is calculated within the Kalman filter using a trilateration algorithm.
The user interface (Fig. 5) of the augmented navigation system guides the diver along
the selected track inside the archeological site. It has been implemented by means of the
cross-platform game engine Unity. Using the tablet position and orientation, the AR tool
shows a 3D map representing the environment around the diver, adding useful
Fig. 5. UI of the AR diving tool
information about underwater artifacts and structures. Moreover, some additional data
are displayed: system status, battery charge, water temperature, and depth. The diver
can also change the view mode or open the camera device to shoot photos that are
automatically geo-referenced.
6 Results
The paper has presented an innovative system that fits the Unesco’s recommendations
for a respectful exploitation of the underwater cultural heritage and represents a concrete
and efficient response to the challenge proposed by the European Parliament for a smart,
sustainable and inclusive growth of the maritime and coastal tourism.
In particular, this system, developed in the VISAS project, integrates virtual and
augmented reality technologies to make more engaging, attractive and educative the
exploitation, both inside and outside, of the underwater archaeological sites. In fact, if
on one side the AR tool improves the diver tourists’ experience in the submerged
archaeological site, on the other the VR tool provides to non-diver tourists a 3D virtual
tour of the underwater cultural heritage.
The VISAS project will end in September 2016 and the remaining activities are
focusing on the experimentation and user studies of the developed tools. The VR diving
tool has been showcased at BPER Optimist European Championship 2016, which has
been held in Crotone, Italy, on July 16–22, 2016 (Fig. 6a). After exposure to the system,
users completed standardized satisfaction questionnaires that confirmed the positive
benefits and added value that VR technologies can bring in the exploitation of under‐
water archaeology sites. About the AR diving tool experimentation activities have been
carried out in the underwater archaeological site of Punta Scifo, located in the East coast
278 F. Bruno et al.
Fig. 6. User interacting with the virtual diving tool (a) and testing the augmented diving tool in
the underwater site of Punta Scifo (b)
of Calabria at 10 km far from Crotone, where raw and semi-finished marble products,
transported by Roman cargo ships, lay on the seabed at a depth of 7 m (Fig. 6b).
Acknowledgements. The VISAS Project (Ref. Start-Up PAC02L2-00040), has been financed
by the MIUR under the PAC Programme. The authors would like to thank the Marine Protected
Area of Capo Rizzuto, the Soprintendenza per i Beni culturali e ambientali del Mare della Sicilia
for the permission to conduct the experimentation in the sites of Cala Minnola and the
Soprintendenza per i Beni Archeologici della Calabria for the site of Punta Scifo.
References
1. Guérin, U.: Underwater cultural heritage and its potentials for sustainable maritime and coastal
tourism. UNESCO Convention on the Protection of the Underwater Cultural Heritage (2001)
2. Unesco 2001: Convention on the protection of the underwater cultural heritage, 2 November
2001. http://www.unesco.org
3. Bruno, F., Gallo, A., Barbieri, L., Muzzupappa, M., Ritacco, G., Lagudi, A., La Russa, M.F.,
Ruffolo, S.A., Crisci, G.M., Ricca, M., Comite, V., Davidde, B., Di Stefano, G., Guida, R.:
The CoMAS project: new materials and tools for improving the in-situ documentation,
restoration and conservation of underwater archaeological remains. Mar. Technol. Soc. J.
50(4), 108–118 (2016)
4. Bruno, F., Muzzupappa, M., Gallo, A., Barbieri, L., Spadafora, F., Galati, D., Petriaggi, B.D.,
Petriaggi, R.: Electromechanical devices for supporting the restoration of underwater
archaeological artefacts. In: Proceedings of IEEE/MTS OCEANS 2015 Conference, Genova
(Italy), 18–21 May 2015
5. Bruno, F., Muzzupappa, M., Lagudi, A., Gallo, A., Spadafora, F., Ritacco, G., Angilica, A.,
Barbieri, L., Di Stefano, G., Guida, R., Di Lecce, N., Saviozzi, G., Laschi, C.: A ROV for
supporting the planned maintenance in underwater archaeological sites. In: Proceedings of
IEEE/MTS OCEANS 2015 Conference, Genova (Italy), 18–21 May 2015
6. Skarlatos, D., Demestiha, S., Kiparissi, S.: An ‘open’ method for 3D modelling and mapping
in underwater archaeological sites. Int. J. Heritage Digit. Era 1(1), 1–24 (2012)
7. Pearson, C.: Conservation of Marine Archaeological Objects. Elsevier, Burlington (2014)
8. Davidde, B.: Underwater archaeological parks: a new perspective and a challenge for
conservation—the Italian panorama. Int. J. Nautical Archaeol. 31(1), 83–88 (2002)
9. European Commission Directorate General for Maritime Affairs and Fisheries: A European
Strategy for more Growth and Jobs in Coastal and Maritime Tourism. Communication from
the commission to the European Parliament, the council, the European and economic and
social committee and the committee of the regions, 20 February 2014
10. Bruno, F., Lagudi, A., Muzzupappa, M., Lupia, M., Cario, G., Barbieri, L., Passaro, S.,
Saggiomo, R.: Project VISAS – virtual and augmented exploitation of Submerged
Archaeological Sites: overview and first results. Mar. Technol. Soc. (MTS) J. 50(4), 119–129
(2016)
11. Barceló, J.A, Forte, M., Sanders D.H.: Virtual Reality in Archaeology. BAR International
Series, vol. 843. Archeopress, Oxford (2000). Edited by Sanders, D.H., Forte, M.
12. Pavlidis, G., Koutsoudis, A., Arnaoutoglou, F., Tsioukas, V., Chamzas, C.: Methods for 3D
digitization of cultural heritage. J. Cult. Heritage 8(1), 93–98 (2007)
13. Bruno, F., Bruno, S., De Sensi, G., Luchi, M.L., Mancuso, S., Muzzupappa, M.: From 3D
reconstruction to virtual reality: a complete methodology for digital archaeological exhibition.
J. Cult. Heritage 11(1), 42–49 (2010)
14. Barbieri, L., Bruno, F., Mollo, F., Muzzupappa, M.: User-centered design of a Virtual Museum
system: a case study. In: Eynard, B., Nigrelli, V., Oliveri, S,M., Peris-Fajarnes. G., Rizzuti,
S. (eds.) Advances on Mechanics, Design Engineering and Manufacturing. LNME, pp. 155–
165. Springer, Heidelberg (2017). doi:10.1007/978-3-319-45781-9_17
15. Almeida, L.: Archeoguide: an augmented reality guide for archaeological sites. IEEE Comput.
Graph. Appl. (2002)
16. Varinlioğlu, G.: Data collection for a virtual museum on the underwater survey at Kaş, Turkey.
Int. J. Nautical Archaeol. 40(1), 182–188 (2011)
17. Chapman, P., Conte, G., Drap, P., Gambogi, P., Gauch, F., Hanke, K., Richards, J.: Venus,
virtual exploration of underwater sites. In: Proceeding of Joint Event CIPA/VAST/EG/Euro-
Med (2006)
18. Haydar, M., Roussel, D., Maïdi, M., Otmane, S., Mallem, M.: Virtual and augmented reality
for cultural computing and heritage: a case study of virtual exploration of underwater
archaeological site. Virtual Reality 15(4), 311–327 (2011)
19. Chapman, P., Roussel, D., Drap, P., Haydar, M.: Virtual exploration of underwater
archaeological sites: visualization and interaction in mixed reality environments, pp. 141–148
(2008)
20. Gallo, A., Angilica, A., Bianco, G., De Filippo, F., Muzzupappa, M., Davidde, B., Bruno, F.:
3D reconstruction and virtual exploration of submerged structures: a case study in the
underwater archaeological site of Baia (Italy). In: The 13th International Symposium on
Virtual Reality, Archaeology and Cultural Heritage VAST (2012)
21. Katsouri, I., Tzanavari, A., Herakleous, K., Poullis, C.: Visualizing and assessing hypotheses
for marine archaeology in a VR CAVE environment. J. Comput. Cult. Heritage (JOCCH)
8(2), 10 (2015)
22. Leinikki, J., Kanerva, J., Panu Oulasvirta, P., Syvärant, J.: Personal underwater tablets bring
new possibilities to data collection, communication, geolocation and diver monitoring. In:
Lang, M.A., Sayer, M.D.J. (eds.) Proceedings of the 2013 AAUS/ESDP Curaçao Joint
International Scientific Diving Symposium, Curaçao, 24–27 October 2013, pp. 163–166.
American Academy of Underwater Sciences, Dauphin Island (2013)
23. Lagudi, A., Bianco, G., Muzzupappa, M., Bruno, F.: An alignment method for the integration
of underwater 3D data captured by a stereovision system and an acoustic camera. Sensors
16(4), 536 (2016)
280 F. Bruno et al.
24. Bruno, F., Lagudi, A., Passaro, S., Saggoimo, R.: Opto-acoustic 3D reconstruction of the
“Punta Scifo D” shipwreck. In: 1st IMEKO TC4 International Workshop on Metrology for
Geotechnics, MetroGeotechnics 2016, pp. 327–333 (2016)
25. Bruno, F., Lagudi, A., Gallo, A., Muzzupappa, M., Petriaggi Davidde, B., Passaro, S.: 3D
documentation of archeological remains in the underwater park of Baiae. Int. Arch.
Photogrammetry Remote Sens. Spat. Inf. Sci. 40(5), 41 (2015)
26. Rende, S.F., Irving, A.D., Bacci, T., Parlagreco, L., Bruno, F., De Filippo, F., Montefalcone,
M., Penna, M., Trabucco, B., Di Mento, R., Cicero, A.M.: Advances in micro-cartography: a
two-dimensional photo mosaicing technique for seagrass monitoring. Estuar. Coast. Shelf Sci.
167, 475–486 (2015)
27. HTC Vive. https://www.htcvive.com
Interacting with Simulated Archaeological Assets
Arian Goren1 ✉ , Kay Kohlmeyer1, Thomas Bremer2, Susanne Brandhorst2,

( )
Arie Kai-Browne1, Felix Balda2, David Strippgen3, and Sebastian Plesch3

1
Landscape Archaeology,
University of Applied Sciences Berlin, Berlin, Germany
[email protected], {Kay.Kohlmeyer,
Arie.Kai-Browne}@HTW-Berlin.de
2
Game Design, University of Applied Sciences Berlin, Berlin, Germany
{Thomas.Bremer,Susanne.Brandhorst,Felix.Balda}@HTW-Berlin.de
3
Computer and Media Sciences, University of Applied Sciences Berlin, Berlin, Germany
{David.Strippgen,Sebastian.Plesch}@HTW-Berlin.de
Abstract. Digital and 3D data are common components in current archaeolog‐

ical work, and expectations regarding their utilization in contextualizing archaeo‐
logical knowledge are steadily on the rise. The rapid progress in real-time
rendering software and more accessible computational power enables integrated
data-sets to (re)gain relevance in the process of interpreting archaeological
contexts. Retaining high level of details and correct geometric relations of a
complex scene while reconciling inherent variations in the scale, format, and
resolution of input data (including 2D legacy data and 3D field recordings) has
been already successfully achieved in the simulation of the Temple of the Storm
God of Aleppo, realized by an interdisciplinary working group in the HTW Berlin.
The current paper addresses the modification of virtual and immersive environ‐
ments within the field of cultural heritage, and evaluating their potential as tools
in interpretative archaeological processes. Based on widely available game tech‐
nology, two applications are presented, supporting real-time interaction and
collaborative work within a single modeled space.
1 Introduction
Incorporating 3D data capturing tools has become a common practice in archaeological

work worldwide, making archaeological data more available for post-processing and
presentation, but also much more voluminous than ever before. Also the array of data
formats, the variety of processing tools, and the range of devices enabling presentation
of digital and digitized data is much wider. The challenge, therefore, modern archaeol‐
ogists face is setting proper standards regarding data-sets manipulation, sophisticated
visualizations, and flexible data application. To meet the rise in expectations, co-oper‐
ating with other researchers or specialists in interdisciplinary frameworks gathers
momentum. This underlines a need to articulate the imperatives for engaging in an
archaeological research, and carefully employ techniques which are relevant to chosen
archaeological analysis and work methodologies. Increased incorporation of advanced
computer and game technologies does not only expand an archaeologist’s toolkit, but
also encourages re-thinking work procedures and work assumptions. In particular,

DOI: 10.1007/978-3-319-48496-9_23
282 A. Goren et al.
virtual environments (VE) and virtual reality (VR) can be utilized to experiment with
new approaches to interpret archaeological assets and provide new means, for both the
public as well as scholars, to re-engage with cultural heritage knowledge. Contrary to
the more common hyper-realistic nature of 3D simulations associated with popular
culture, when purposed for scientific use it is essential to maintain the correct relations
between different data-sets and varying resolutions involved in assembling archaeolog‐
ical information systems.
A recently developed application evaluates the potential of VR platforms and immer‐
sive devices as ground-braking tools in interpretative archaeological research. The
application allows for real-time interaction with a modeled space and its assets, and
supports remote collaborative work within a single simulated space. Demonstrated on
the game engine generated 3D simulation of the Temple of the Storm God in Aleppo,
purposing immersive devices to the field of cultural heritage is explored.
2 Implementing and Manipulating Integrated Data in Virtual

Environments
A viable 3D simulation of the Temple of the Storm God and its assets, which correctly
visualize the various temple components, was constructed through workflows compiled
by members of a dedicated interdisciplinary working group [1]. These workflows
describe the steps taken to properly convert the various inputs comprising the virtual
scene. The datasets used to compile the virtual scene include data captured originally in
3D in the temple compound itself in 2010, available satellite imagery, and 2D legacy
data. Prior to compiling the virtual scene, processing these varied inputs for further
manipulation followed specific pipelines in order to meet the project objectives and
while taking methodological concerns into account [1]. The resulting product is a
detailed and multi-scalar simulation of the temple compound with its architectural
elements and finely carved inscriptions. The geometrical relations of the objects were
accurately maintained despite rigorous file size reduction, achieved by implementing
decimation algorithms [1–4]. Using capabilities offered by media design and game
technology, a 3D simulation of the temple was embedded in a VE, facilitated with inter‐
active functionality and displayable on both desktop as well as immersive devices, such
as head mounted displays (HMD).
Several aspects need to be considered when implementing 3D scientific data in VR
platforms. Technical issues might be the most obvious ones. From our experience with
several game engines, achieving high representational fidelity is much more a matter of
the engine’s architecture and the quality and Level of Details (LoD) it supports rather
than the quality and accuracy of the fed data. As programming a costume game engine
is not part of the project goals, and in order to keep costs low, a strategic decision was
taken to generate the simulation on available gaming platforms (several cost-free plat‐
forms available for non-commercial use were tested. In this paper the Unreal Engine 4
is referred to). Formerly, relying on cost-free development kits restricted the range of
available game engines to those offering only limited access to code-editing and asset
manipulation. This approach substantially constrains the ability to simulate complex
Interacting with Simulated Archaeological Assets 283
scenes comprising multiple assets with a wide range of LoDs. Fortunately, game tech‐
nology is a rapidly developing industry. Newer versions of game engines are regularly
released to the market, implementing better rendering algorithms, deeper development
kits, and higher representational fidelity. Once the groundwork is set and the required
3D data is processed, converting the simulated scene to another game engine or a newer
version is relatively simple, without apparent loss of details or data quality. Nonetheless,
this step should be supported with hardware components which correspond to the
system’s requirements.
Technical issues are only the backstage to other fundamental concerns regarding
content and functionality. The potential affordances of a certain VE are prompted
through several factors, determine the user’s experience and perception of the environ‐
ment. Such factors can be, for example, the scene’s design, smoothness of display, choice
of interface, or type and range of user allocated actions [5, 6]. In the developed appli‐
cations presented here, navigational tools were implemented, meant to enhance an
exploratory affordance and help the user to develop a better sense of presence and spatial
acquaintance with the modeled environment. It is important to note the difference
between a sense of presence, or embodiment - the reaction a user might develop while
experiencing a VE - and the notion of spatial immersion, which can be described as the
constellation of components and characteristics comprising a certain VE [6]. While the
latter depends on system architecture and hardware configuration, the former has to do
with the content and actions embedded in the VE. Whether an immersive system will
encourage a user to develop a sense of presence depends on the system’s configuration
and the representational fidelity of the environment. However, whether a user will be
motivated to engage with the VE and the assets in it has much to do with the affordances
offered by the VE. An effective immersive system should convince us, that is our
perception, in the real-life properties of the virtual simulation we are experiencing.
The challenges in reconciling between the different methodologies of game tech‐
nology and archaeology are made more clear when trying to adjust game engines to
produce complex multi-scalar and visually versatile virtual simulations. By their nature,
game engines are constructed to deliver first-person experience and are optimal for
presenting vivid representations of a (virtual) world from a subjective point of view.
Hence, some game engines offer better rendering when the simulated scenes are spatially
limited. Although it is usually possible to depict endless simulated plains, distant objects
will be often simplified as a “backdrop” to the main scene.
3 Virtually Re-contextualizing the Temple of the Storm God
The created VE depicts a 3D simulation of the Temple of the Storm God in Aleppo, a
compound measuring 42 × 42 m with an interior cella of approximately 27 × 17 m,
dated to the 1st half of the 2nd millennium BCE [7]. The modeling of the temple was
created with common 3D processing software [1], rendered in the Unreal Engine. The
resulting simulation retains the temple’s floor plan in several scales while maintaining
accurate geometric relations of the architectural elements (Fig. 1). A key aspect in
284 A. Goren et al.
modeling the temple is the ability to distinguish various spatial levels, such as very fine
chisel marks and hieroglyphic inscriptions from remote mountain ranges.
Fig. 1. The virtual simulation of the Temple of the Storm God compound rendered in the Unreal
Engine 4.
Two types of applications were tested in the virtual simulation, both well established
in the world of game technology and immersive environments. One of the applications
uses an HMD. By nature, such devices are designed to generate a highly immersive
environment, which can induce users to very vivid reactions and a strong sense of pres‐
ence, mimicking an illusory sense of body ownership [8–10 and references therein].
The other application refers to Multiplayer Computer (Online) Games (MCG/MOG),
among the most widely distributed types of popular entertainment. Research on online
game addiction [11, 12] indicates the profound immersive quality MOGs can have, even
though - contrary to games played using immersive devices such as HMDs - for the most
part they are played in front of a computer screen. Research on the educational and
sociological aspects of computer and online games [13–15] reveals that such games
provide users valuable opportunities to engage in social interactions. MCGs and MOGs
are powerful platforms, motivating participants to fulfill effective learning tasks and
encouraging them to explore, communicate, and collaborate.
4 Interactively Exploring the Temple of the Storm God
Having user experience in mind, this application implements an Oculus Rift DK2 HMD
device to the virtual simulation of the temple compound (Fig. 2), with free movement
in the modeled space, and direct access to interaction being emphasized. Enabling these
actions aims to facilitate spatial affordances - high degree of acquaintance with the
temple compound and its layout, and consistent assessment of scales and measures from
different locations within the compound. However, a learning affordance was also
assessed through enabling the interaction with archaeological content, in this case the
temple’s decorated reliefs.
Fig. 2. The hardware setup used in the HMD application implemented in the virtual environment
of the temple. The user wears an Oculus Rift DK2 HMD device. Movement in the virtual
environment is controlled through a hand held XBox 360 controller. The user’s movements are
captured through an external Oculus positional tracking camera positioned in front of the user.
Nonetheless, several limitations and challenges are inevitable when using immersive
devices such as HMDs, relating to representation of the virtual avatar and to navigation
in the VE. Motion sickness (or simulator sickness) is the most prominent challenge to
be addressed in this context. A variety of factors could trigger this phenomenon. Some
refer to the construction of the system, such as frame rate, navigation speed, or display
characteristics. Other factors are rather subjective and depend on individual circum‐
stances, such as age, previous experience with VE systems, or medical condition [16–
18]. The most common occurrence of motion sickness are instances of spatial disori‐
entation or general discomfort, triggered by the effects that an immersive experience has
on the proprioceptive system. Movement is one of the most obvious examples of motion
sickness: while the physical body of a user is stationary in the real world (s)he experi‐
ences its virtual representation, or virtual avatar, as moving or flying without any corre‐
sponding vestibular or kinesthetic cues.
A possible way to reduce some of these disorienting effects is to bridge the gap
between the physical gestures of a user and those of the virtual counterpart. A solution
implemented in the tested HMD application was to allow the user to control navigation
in the modeled space with a hand held controller (here an XBox 360 joystick). However,
in order to do so, first the movements of the user need to be captured and mapped on to
the gestures of the avatar in the VE. A basic HMD configuration positions only a single
tracking device. The tracking camera provided with the Oculus Rift device, which was
used as a default for the temple’s virtual scene (Fig. 2), tracks only the position of the
HMD and not that of the full body. However, movement mapping does not need to be
perfect, suffice it when gestures of movement or directions are generally simulated in
order for a user to develop a sense of presence and experience embodiment [6].
286 A. Goren et al.
Stepping out of the range of the tracking device pauses the motion tracking and the
display fades to black. This pre-defined characteristic could also induce motion sickness.
But more importantly, this action deeply effects the user’s movement, in the physical
world as well as in the virtual space. Extending the tracking volume (of both the user
and the avatar) can be rather easily achieved with additional devices/detectors, or by
switching to technologies using motion capture systems or magnetic based tracking
(trading-off a light-weight single laptop and HMD with a much less portable system).
However, even if systems are switched or tracking devices are added, the user will still
be literally confined to staying on ground level. Performing a flight, for example in order
to observe objects from above, will still, in most likelihood, induce motion sickness,
since it can be simulated virtually but not tracked physically.
Reconciling between system mobility and limited tracking volume was settled in the
developed application through implementing a virtual dynamic platform, on which the
avatar is placed. In fact, this solution improves user experience in a twofold way. On
the one hand, the tracking volume is substantially extended. The user is granted with an
almost unlimited range of movement in the simulated temple, and so the risk of stepping
out of the scene is reduced. In addition, also motion sickness is reduced by overcoming
the discrepancy between the physical movement and its simulation in the temple.
Fig. 3. A virtual movable dynamic platform and visualization of tracking volume boundaries.
(a) The user’s representation in the VE (as an avatar) is visualized on the dynamic platform, while
the user’s body (here slightly crouching) is approximated from the position of the HMD respective
the tracking device. (b) Schematic depiction of the tracking volume covered by the tracking device
positioned in front of the user. (c) The walls of the confining box blend in front of the user/the
avatar when approaching the limits of the tracking volume, signaling the need to re-position.
The virtual platform, designed to resemble a “flying carpet” (Fig. 3a), is visible to
the user and semi-transparent to not distract the view. The platform’s position in the VE
is manipulated by the user through a hand held controller, allowing to actively explore
the entire modeled temple and approach objects from multiple angles.
The platform is structured as a confining box with dynamically changing walls. In a
static position, only a subset of the platform is visible at the feet of the user (as seen in
Fig. 3a). When the limits of the tracking volume are approached (Fig. 3b), a colored
wall appears in front of the user, signaling a need to reposition (Fig. 3c). Moving the
dynamic platform results in the re-positioning of the tracking volume within the virtual
space, and does not require to re-position the avatar (essentially negating the need to
physically move in the real world). The confining box becomes visible also when the
platform is shifted in the virtual space. The user thus can have a better spatial reference
and can apprehend that the platform (or avatar) is what being moved rather than its own
physical body.
5 Remote Collaboration in a Shared Virtual Space
Another tested application builds on the concept of multi-user systems, or multiplayer

gaming platforms. Whether played online or implemented in VEs, multi-user systems
demonstrate powerful immersive characteristics. Such systems can support basic
features, such as location, orientation, agency, and communication [9, 19]. These
features are capable of inducing a strong sense of embodiment and participation, albeit
the fundamental difference in their configuration compared to other immersive VEs,
such as HMDs or DIVE/CAVE systems [13, 20].
Running on the same virtual simulation of the temple, the system is configured to
support a first-person point of view and the scene is displayed on a screen. Similar to
the HMD application, the temple compound can be actively explored and the different
assets can be approached from up close. Also the avatar movements and parts of the
virtual body representation are visible to the user, encouraging a sense of orientation
and direction in the VE. The main difference compared to the user-oriented HMD appli‐
cation, is the design of this tested multi-user system, which aims to enhance producer
expertise. In order to reach that goal an innovative approach was taken, maximizing
real-time interaction and knowledge sharing between peers using common interfaces,
such as desktop PC.
The concept of the application’s design follows two key objectives: real-time
communication and interactive cooperation, both are essential properties in conducting
meaningful collaborations in VEs [19, 20]. The platform offers participants to carry out
remote work sessions, thus supporting knowledge transfer between professionals and
decision makers. Using the application, multiple users can conduct remote work sessions
in a faster and more efficient way while sitting in front of their personal computer screens.
During a work session, participants can operate two main types of user-asset inter‐
action. They can open information pop-ups (Fig. 4), containing information regarding
the characteristics of a certain object. Additionally, assets in the virtual scene can be
288 A. Goren et al.
edited and manipulated on-the-fly (Fig. 5a), while these and other actions are evaluated
in real time through verbal communication.
Fig. 4. Informational pop-ups regarding the properties of specific objects can be activated by the
user when approaching an object.
Fig. 5. The virtual simulation of the Temple of the Storm God as seen from the point of view of
a single participant in the multi-user application. (a) A modeled wall added to the virtual scene as
part of testing a suggested reconstruction of the pedestal wall in the Temple’s compound. (b) The
avatar of another participant is seen while inspecting the cult images of the Storm God and King
Taita of Philistin to his right.
The temple compound can be navigated and explored simultaneously by multiple

participants (Fig. 5b). The properties of existing objects (for example its spatial position
or physical measures) can be edited, and new objects can be added to the scene, however
only one object can be manipulated by a single user at a time. While an editing action
takes place, the object is highlighted so other peer(s) can be aware of changes made to
the scene. Object manipulation in the application is based on an underlying database,
containing the model and its assets. Establishing a direct link to the underlying database
allows to access assets in real-time without the need to recompile the entire application
after each editing task.
Enabling an advanced asset-interaction is particularly important in order to organize
and carry out tasks successfully in a shared virtual space. The developed user-asset
interactions specifically aim to open up new approaches to re-contextualize cultural
heritage knowledge, for example when testing assumptions regarding architectural
reconstructions of both diachronic and synchronic implications. However, much as in
real-life, peer communication and collaboration are essential to support a sense of co-
presence in VE, and can be decisive factors in carrying out joint tasks successfully [19,
20]. Hence, both written and verbal communication options are offered in the applica‐
tion, as embedded text messaging or via VoIP technology offered by third-party services.
Different from the HMD application, some prerequisites need to be met in order to
operate the multi-user system. Participating in a session requires internet accessibility
in order to connect to the server hosting the VE and the temple’s simulation. Hardware
configuration (in particular graphic card performance) as well as system and network
stability (which are dependent on the available client/server infrastructure) might also
present limiting factors to carrying out a successful collaborative session. At the
moment, the design of the application requires all participants to load the same model
version on their respective PC in order to share the same simulated space. Designing a
peer-to-peer (P2P) configuration and connecting to a shared asset database can improve
some of these issues and provide better application flexibility and increased stability.
Further aspects that can be improved in future versions of the application relate to phys‐
ical properties and sense of embodiment in the VE. These, however, are more so
dependent on the state of game technology available in the market. Developments in the
field are anticipated in the coming future with the introduction of advanced tactile user-
interfaces and newer versions of end-user immersive devices.
6 Closing Remarks
Taking advantage of capabilities offered by media design and game technology, dedi‐
cated applications were implemented on a 3D simulation of the Temple of the Storm
God in Aleppo. The 3D model of the temple compound and its assets are generated in
a game engine, and based on data integrated from diverse sources with varying scales.
The purpose of the presented applications is to allow real-time interaction with the virtual
temple space and its assets. While an HMD application emphasizes user experience, a
multi-user system addresses scientist and professionals in the fields of archaeology and
cultural heritage. In particular, the multi-user system enables conducting a remote
collaborative work within a single shared space. Both applications facilitate interactive
functionalities and are displayable on PC desktop as well as immersive devices. In
290 A. Goren et al.
general, virtual environments can be very effective tools in inducing a strong sense of
presence and embodiment, and motivate engagement with individual as well as joint
tasks. Both applications implement further advanced functions which allow to directly
manipulate and edit assets in real-time. The work of an interdisciplinary group of
archaeologists, media designers, and computer scientists presented in the paper lays
ground for meaningful remote collaborative work, where peers can act and communicate
in the same virtual space. Such capabilities are particular interest also for professionals
seeking to engage in re-thinking work procedures as well as re-contextualizing cultural
heritage knowledge. With future developments in game technology and newer versions
of immersive devices released to the market, soft-spots such as motion sickness and
tactile feedback are expected to be improved.
Acknowledgements. Former sponsors of the project and related fieldwork are: the IFAF Institute
Berlin, the World Monuments Fund, the German Research Foundation, the Gerda Henkel
Foundation, TOPOI Cluster of Excellence at the Free University of Berlin.
References
1. Goren, A., Kohlmeyer, K., Bremer, T., Kai-Browne, A., Bebermeier, W., Öztürk, D., Öztürk,
S., Müller, T.: The virtual archaeology project – towards an interactive multi-scalar 3D
visualisation in computer game engines. In: Traviglia, A. (ed.) Across Space and Time,
Selected Papers from the 41st Computer Applications and Quantitative Methods in
Archaeology Conference, Perth, WA, 25–28 March 2013, pp. 386–400. Amsterdam
University Press (2015)
2. Alliez, P., Ucelli, G., Gotsman, C., Attene, M.: Recent advances in remeshing surfaces. In:
De Floriani, L., Spangnulo, M. (eds.) Shape Analysis and Structuring, pp. 53–82. Springer,
Heidelberg (2008)
3. Merlo, A., Dalcò, L., Fantini, F. Game engine for cultural heritage: new opportunities in the
relation between simplified models and database. In: Guidi, G., Addison, A.C. (eds.) 18th
International Conference on Virtual Systems and Multimedia (VSMM), Proceedings of the
VSMM 2012 Virtual Systems in the Information Society, Milan, Italy, 2–5 September 2012,
pp. 623–628. Institute of Electrical and Electronics Engineers (IEEE), Piscataway (2012)
4. Merlo, A., Sánchez Belenguer, C., Vendrell Vidal, E., Fantini, F, Alipetra, A.: 3D model
visualization enhancements in real-time game engines. In: Bohem, J., Remondion, F., Kersten,
T., Fuse, T., Gonzalez-Aguilera, D. (eds.) 5th International Workshop 3D-Arch 2013: Virtual
Reconstruction and Visuzlization of Complex Architectures, Trento, Italy, 25–26 February
2013, vol. XL-5/W1, pp. 181–188. International Archives of the Photogrammetry, Remote
Sensing and Spatial Information Sciences (ISPRS) (2013)
5. Dalgarno, B., Lee, M.J.: What are the learning affordances of 3-D virtual environments? Br.
J. Educ. Technol. 41(1), 10–32 (2010)
6. Slater, M.: A note on presence terminology. Presence Connect 3(3), 1–5 (2003)
7. Kohlmeyer, K.: Der Tempel des Wettergottes von Aleppo. Baugeschichte und Bautyp,
räumliche Bezüge, Inventar und bildliche Ausstattung. In: Kamlah, J. (ed.) Temple Building
and Temple Cult, pp. 55–78. Harrassowitz Verlag, Wiesbaden (2012)
8. Ehrsson, H.H.: The experimental induction of out-of-body experiences. Science 317(5841),
1048 (2007)
9. Kilteni, K., Groten, R., Slater, M.: The sense of embodiment in virtual reality. Presence
Teleop. Virtual Environ. 21(4), 373–387 (2012)
10. Slater, M., Perez-Marcos, D., Ehrsson, H.H., Sanchez-Vives, M.V.: Inducing illusory
ownership of a virtual body. Front. Neurosci. 3(2), 214–220 (2009)
11. Kuss, D.J., Griffiths, M.D.: Internet gaming addiction: a systematic review of empirical
research. Int. J. Mental Health Addict. 10(2), 278–296 (2012)
12. Lee, Z.W.Y., Cheung, C.M.K., Chan, T.K.H.: Massively multiplayer online game addiction:
instrument development and validation. Inf. Manag. 52(4), 413–430 (2015)
13. De Freitas, S.: Learning in Immersive Worlds. Joint Information Systems Committee, London
(2006)
14. Steinkuehler, C.A.: Massively multiplayer online games. Paper Presented at the Proceedings
of the 6th International Conference on Learning Sciences: ICLS 2004: Embracing Diversity
in The Learning Sciences, University of California, Los Angeles, Santa Monica, CA, 22–26
June 2004 (2004)
15. Steinkuehler, C.A., Williams, D.: Where everybody knows your (screen) name: online games
as “Third Places”. J. Comput. Mediated Commun. 11(4), 885–909 (2006)
16. Kolasinski, E.M.: Simulator Sickmess in Virtual Environments. DTIC, U.S. Army Research
Unit (1995)
17. Moss, J.D., Muth, E.R.: Characteristics of head-mounted displays and their effects on
simulator sickness. Hum. Fact. J. Hum. Facto. Ergon. Soc. 53(3), 308–319 (2011)
18. So, R.H.Y., Lo, W.T., Ho, A.T.K.: Effects of navigation speed on motion sickness caused by
an immersive virtual environment. Hum. Fact. J. Hum. Facto. Ergon. Soc. 43(3), 452–461
(2001)
19. Mortensen, J., Vinayagamoorthy, V., Slater, M., Steed, A., Lok, B., Whitton, M.:
Collaboration in tele-immersive environments. In: Stürzlinger, W., Müller, S. (eds.)
Proceedings of the Eighth Eurographics Workshop on Virtual Environments 2002.
Eurographics Association, Aire-la-Ville (2002)
20. Benford, S., Bowers, J., Fahlen, L.E., Greenhalgh, C., Snowdon, D.: User embodiment in
collaborative virtual environments. In: Proceedings of the SIGCHI Conference on Human
Factors in Computing Systems, pp. 242–249. Addison-Wesley Publishing Co. (1995)
Virtual Reconstruction 3.0: New Approach
of Web-based Visualisation and Documentation
of Lost Cultural Heritage
Daniel Dworak1,2(B) and Piotr Kuroczyński3

1
Institute of Information Technology, Lodz University of Technology, L ódź, Poland
[email protected]
2
Center for Media and Interactivity, Justus Liebig University Giessen,
Giessen, Germany
3
Herder Institute for Historical Research on East Central Europe,
Marburg, Germany
[email protected]
Abstract. The paper presents the project entitled “Virtual Reconstruc-

tions in Transnational Research Environments the Portal: ‘Palaces and
Parks in former East Prussia”’ in the light of the Semantic Web and Open
Source technologies. The researches are focused on certification, classifi-
cation, annotation, storage and visualisation of 3D data sets, proposing
methodology of the computer-based 3D computer reconstruction of Cul-
tural Heritage, which are still lacking. The multinational and interdisci-
plinary project with interactive 3D models being part of a semantic data
model, is concerned with designing a Virtual Research Environment. Our
approach affects the entire process of digital 3D reconstruction with the
development of an XML schema called Cultural Heritage Markup Lan-
guage as a groundwork for an application ontology.
The results bring new insights into areas such as effective data acqui-
sition, documentation, semantic 3D modelling and visualisation and data
management. They may be useful for the creation of Virtual Environ-
ments and other forms of Cultural Heritage’s interactive presentation
that employ open source visualisations standards (e.g. WebGL technol-
ogy). An aspect that needs to be improved concerns coding and upload-
ing large 3D data sets using alternative formats, with an emphasis on
art and architectural models. We have developed a technique for coding,
long-term storing and decoding 3D geometrical data in 2D PNG files,
which are characterised by small size and lossless compression.
1 Introduction
The art and architectural representation, as well as the research in the field
of art history, are traditionally based on iconographic sources and 3D models.
The development of the Computer Aided Design software accompanied by the
potential growth of the counting performance of the computers led to increased

DOI: 10.1007/978-3-319-48496-9 24
Virtual Reconstruction 3.0 293
application of so called Virtual Reconstruction in research projects and for dis-

semination purposes in the Cultural Heritage sector, in particular museums exhi-
bitions.
Since 1990s we have observed an increased application of Virtual Recon-
struction based on the enthusiasm for innovative visualisation possibilities and
(hyper)realistic representations (renderings) of destroyed or never realised art an
architecture. One of the early academic centres for the Virtual Reconstruction
since the beginning of 1990s was established by Prof. Manfred Koob (2011) at
the Darmstadt University of Technology (Faculty of Architecture). The projects
reflect the development and application of the digital 3D reconstruction through-
out two decades [13]. Since then all the involved researchers on the one hand have
recognized the potentials of the new “holistic approach” in the data processing,
on the other hand they have pointed out the drawbacks of (hyper)realistic visu-
alisation and the lack of documentations standards [4]. The treachery of images
highlights the gap between the image, language and the meaning.
In general, the result of a Virtual Reconstruction is an animation (render-
ings) the recipients are confronted with. The images suggest an imaginary pic-
ture of lasting effect in the sense of “A picture says more than 1000 words!”.
The unsolved questions how to record the interpretative creation process of the
model and preserve a scholarly approved information lead to a critical atti-
tude towards digital 3D models. Beside internationally-recognised principles for
the use of computer-based visualisation by researchers, educators and cultural
heritage organisations, like London Charter, there is no documentation, visuali-
sation and preservation standard for digital 3D models in the above-mentioned
scientific context [3].
Besides, the museums are currently confronted with the challenges of the
mass digitisation of the 3D artefacts (collections), the scientific approach and
digital preservation. There are several documentation standards for digital 3D
models resulting in data models. The XML metadata schemas, like CARARE
2.0, captures the machine-driven process of 3D scanning and the context of the
artefact, and enable rudimentary publishing to digital libraries, e.g. Europeana
[1]. The digital 3D reconstruction (Virtual Reconstruction) differs from the 3D
digitisation, because normally there are no remains that could be scanned. The
digital 3D reconstruction refers to heterogenous sources and is highly interpre-
tative (human-driven). The London Charter introduces the “paradata” (data
about the creative processes of human interpretation and 3D modelling) to cap-
ture the essential process of the digital 3D reconstruction.
To ensure the academic acceptance of digital 3D reconstruction as schol-
arly productions, contributing “comprehensive historical digital 3D models”, a
customised data model is of outstanding importance. Collecting, transforming
and managing the data as central part of the digital 3D reconstruction force
us to consider the formal structure of the data. Otherwise we risk to lose the
data that leads to the final result the digital 3D representation of the research
issue. In a row leaving the 3D visualisation (renderings) stand alone without
context and without scientifically required “digital footnotes.” Additionally,
294 D. Dworak and P. Kuroczyński
the long-term storage of the information and a scholarly approved visualisation

of the 3D models plays a significant role for the acceptance.
2 Designing a Virtual Research Environment for Digital

3D Reconstructions
The ongoing project “Virtual reconstructions in transnational research environ-

ments the web portal ‘Palaces and Parks in former East Prussia”’ [9] focuses on
the challenges concerned with the data processing, documentation and visuali-
sation. The core aim of the project is the design of a data model in reference
to existing documentation standards and the development of a Virtual Research
Environment (VRE) for spatial research projects [7]. As the groundwork serves
the “Cultural Heritage Markup Language” (CHML), a data model in XML-
format specified for the needs of a digital 3D reconstruction [6]. The implemen-
tation and further development of the data model in E-CRM/OWL DL results
in an application ontology, compliant with the ISO 21127:2006 (CIDOC CRM)
[5]. The current state of the CHML application ontology.
The collaborative VRE is set up upon the “Scientific Communication
Infrastructure” (WissKI) [10]. The innovative approach of WissKI is the mar-
riage between the open-source Content Management System (Drupal) and the
graph database with a triple-store based on E-CRM/OWL DL. The system
enables the usage of easy to handle field-based entries and free-text annota-
tions. All entries create RDF-triples following the Semantic Web requirements.
The free-text annotation allows the semantic keywording. The project processed
data is Linked Data.
Fig. 1. Graphical abstract of the project (Color figure online)

The VRE enables the involved, multidisciplinary research team to acquire

data sets and to semantically enrich the records, 3D models included. While
recording and editing the data, the system semi-automatically expands the
graph database, interlinking the data with internal records of the database and
with external Linked Data addressed by the SPARQL Endpoint. The graphical
abstract reveals the centralised backbone of the project, represented by the red
background (Fig. 1).
The activities of the domain expert (e.g. art historian) evaluate the sources
and identify the objects for further 3D reconstruction. Activities, sources and
objects are embedded and annotated within the VRE. The 3D modeller (e.g.
architect) interprets the registered information and creates the digital 3D recon-
struction. Ideally, in result, every identified object includes an interpretative
(hypothetical) 3D representation. The creative provenance of the digital 3D
model (paradata) is annotated by the architect during the modelling process.
All field-based entries and the semantically enriched free-text annotations within
the VRE results in a graph data. The integration of 3D scenes by queries take
place in the prototype of the “Virtual Museum” (VM), an immersive and inter-
active web-based 3D environment (WebGL). The VM combines the semantic
data model with the 3D visualisation providing transparency and validity of
information.
In the following the project-based approach for the web-based 3D visualisa-
tion as well the customised integration of semantically enriched and scholarly
approved 3D models will be explained from the technological point of view.
3 Challenges for 3D Visualisation

The beginnings of 3D graphics on the Internet dates back to the mid-90s when it
began to exploit the potential of graphics accelerators, also in Web browsers. In
2005, there has also been developed Adobe Papervision3D, which revolutionized
the real-time 3D graphics on the Internet. Further technological development
occured with the publication of the latest implementation of the HTML5 stan-
dard. It introduced the possibility to draw graphics using scripts (mostly using
Javascipt) on the < canvas > element simply called canvas. Consortium Khronos
Group (including Intel, NVIDIA, Google, Sony, Apple, Samsung, IBM) proposed
a completely innovative project, developing the extension of JavaScript that have
an access to the 3D API called WebGL [14] with usage of Web browser.
WebGL uses the OpenGL ES 2.0 instructions and allows to view three-
dimensional elements using the element canvas. In order to ensure smooth anima-
tion, WebGL provides hardware supported rendering using graphics card (GPU),
and in the case of hardware incompatibility - rendering on a CPU. Currently,
WebGL is supported by all the latest desktop browsers, even mobile ones. There
are many libraries using the capabilities of WebGL (GLGE, C3DL, Copperlicht,
SpiderGL, SceneJS or Three.js), without installing additional software or plugins
- it is only needed to install a supported browser.
3.1 Storing 3D Geometry
There are formats that are optimized and supposed to be more efficient in Web
appliance (Table 1). Many Internet technologies support COLLADA (DAE) [2]
format because of variety of information that can be stored in one file. This
XML-based format is also supported by various software like the Maya, the 3ds
Max, the SketchUP or the Blender 3D. This format can be sufficient for the
models with small number of vertices, but according to our test model, .DAE
was the biggest one with 480 MB file’s size. The JSON format (JavaScript Object
Notation) is an open standard format with user-friendly notation (e.g. readable
text) and is used to transmit data from server to client easily. It allows to store
huge data sets like vertices, normals, texture coordinates, faces and textures’
names. Although the Web technologies decode JSON files faster than OBJ, due
to specific way of storing geometry with the similar size as OBJ format.
Table 1. Size comparison for different file formats - 1.5 M and 76,625 vertices geometry.
Vertices DAE FBX WRL OBJ JSON

1.5 M 480 MB 41.3 MB 176 MB 267 MB 230 MB
76,625 19.80 MB 3.10 MB 8.28 MB 6.95 MB 6.55 MB
There are also challenges with users hardware, platforms, browsers and com-
puting power. The main problem for 3D graphics in the Web are huge 3D data
sets, which contain 3D models, in particular art and architectural models, terrain
and vegetation models. Transmission of such data via network requires wideband
Internet connection and transmission formats. There are standarized transmis-
sion formats for Audio (MP3), Video (H.264), and images (JPEG, PNG) but
no transmission format exists for 3D computer graphics yet [12]. Popular 3D
data formats like FBX, OBJ, WRL, COLLADA (DAE), DWG are too heavy to
send them via Internet in real time. An aspect that needs to be improved con-
cerns coding and uploading large 3D data sets using alternative formats, with
an emphasis on art and architectural models. First steps were aimed to the way
of saving and storing 3D models data. This problem is being tried to be solved
by Khronos Organization [12] using content pipeline: Collada − > glTF − >
Open GL, OpenGL ES, or WebGL. A glTF asset is represented by JSON file (for
node hierarchy, materials, and cameras), binary files (geometry and animations),
JPEG and PNG image files (textures), and GLSL text file (shaders). Another
idea, that seemed to be worth considering is coding 3D data sets to 2D binary
PNG files. We have developed a technique for storing 3D geometrical data in
2D PNG files, which are characterised by small sizes and lossless compression
(Fig. 2). Additionally, the PNG compression can be used for long-term preser-
vation purposes, due to its portable and standardised form. Our research has
shown that the 3D data encoding in PNG file reduces its size and time needed
to transfer them via Internet.
Fig. 2. (1) Prototype of the Virtual Museum: elements of geometry - vertices, faces,
normal and UV’s, (2) structure of PNG file with geometry (vertices, faces, normals,
UV’s and materials) and (3) PNG output file with geometry for testing model of
fireplace.
Basically, there are five types of data to define 3D model: vertices (space
coordinates x, y and z), faces (set of vertices’ indices which creates single poly-
gon), normals (vectors of surface’s orientation for lighting calculations), texture’s
coordinates (called UV’s), material properties (like color coefficients and images
for texture). It have been decided to use JSON (JavaScript Object Notation)
format for storing 3D data. This format is an open standard with user-friendly
notation (e.g. readable text) and is used to transmit data from server to client
easily. The conception of conversion JSON to PNG files is based on saving any
3D data to RGB channels of 2D image (Fig. 2). Then, every value (vertex, nor-
mal, UV’s) is splitted into integer and fractional parts and stored in R, G, B
channels. The Alpha channel (A) identifies kind of data: vertex (A = 128), faces
(A = 100), normal (A = 50), texture coordinate (A = 25), material (A = 255).
At the end of PNG file, information about materials are saved. There has been
proposed JSON based sequence of properties like: materials’ name, colors, tex-
tures’ names or transparency. A textual information (for example the name of
a texture map) is stored in ASCII code - one character in single pixel’s channel
but material’s colors - as RGB pixels.
We have decided to use PNG file, because it supports lossless data com-
pression and was designed for transferring it over the Internet. The compression
algorithm implemented for PNG file is called DEFLATE and it builds a data dic-
tionary of information occurring in an original stream. This method eliminates
redundant data, because, when any part of data occurs once, it may be used
many times. This is an important point in our research, because many art and
architectural models are symmetrical and tend to have a lot of common vertices,
faces and other characteristics. This approach is likely to help to decrease the
size of files (Table 2). For example, in case of raw 3D data with 550,000 vertices
stored in OBJ we have 79 MB file. After saving it as PNG we only have approxi-
mately 11.5 MB. There is also an improvement for small objects (35,000 vertices)
OBJ file was about 3.5 MB, and a PNG file was only 0.76 MB (780 KB). More-
over, modern web technologies are optimised for streaming pictures and movies.
This reduces the uploading time of a PNG file, compared to OBJ file.
Table 2. Files’ size comparison for architectural models, according to OBJ and PNG
format.
Model No. vertices OBJ [KB] PNG [KB] Ratio OBJ:PNG

Portrait 11 439 2 124 518 4,1
Chairs 12 618 2 098 462 4,54
Column 17 901 3 083 709 4,35
Sala001 24 958 3 727 807 4,62
Sala002 28 480 12 971 2 334 5,56
Sala112 39 202 6 882 1 370 5,02
Stairways 44 525 9 246 1 372 6,74
It has been proposed to use a GPU rendering pipeline, supported by WebGL

technology, to decode large two-dimensional data sets to tridimensional geom-
etry. Our researches revealed that a model with 77,407 vertices saved in OBJ
file (13 MB) downloads about 25 s, while model encoded by authors’ algorithm
in PNG (2.25 MB) only 4.5 s. For the same PNG file, a period of time that is
needed to decode a data set is also short (less than 1 s), but GPU decoding is
nine times faster than CPU due to calculation’s parallelism.
4 Integration of 3D Content Within WebGL and WissKI

At the beginning of the project, there has been proposed ways of storing, placing
and giving names for 3D models. The whole scene is splitted into smaller parts,
familiar to the hierarchical structure of the spatial artefacts and architectural
ensembles and objects. It means the “Manor House” hasPart “Facades”, “Roof”,
“Storages”, etc. The storage (e.g.: 1. floor) hasPart “room no. 1” hasPart “Wall”
hasPart “Fireplace” hasPart “Portrait A. And the other way around: the “Fire-
place” is a part of the “Wall”. It makes it possible to store, annotate and retrieve
information about every model apart. What is more, there is also an opportu-
nity to receive data about “Fireplace” while asking only about “Wall” as it is a
part of it. The challenge occurs in the final scene’s creation while having many
separate 3D models. The origin point (0, 0, 0) is set to the entrance of the Palace.
Then, each 3D model is put in the right place, according to the basic model of
facade, which makes it fast and simple to build the scene (with every model).
Every 3D model is stored in triplestore database as the result of Activity
event and the “3D reconstruction” process, for example modelling one. It means
that every object has its own and unique URI with accesible website. It pro-
vides the way of describing every needed aspect, like topic, title set (in different
languages), timespan, actor, sources used, additional events (creation, modifica-
tion or destruction), rich text annotations and much more (Fig. 3). The file with
geometry is attached by modeller and placed in Geometry field.
The customized WissKI system also provides a way of storing, describing
and displaying data saved in a variety of modern formats. They are accesible
Fig. 3. The element “M odellingEvent” within the object view and the RDF-triples of
“M odelling Event”
via WebGL based canvas: for 2D files (JPEG, TIFF, PNG) and 3D (OBJ, DAE,
JSON, 3dPNG) as well. While creating the 3D scene, the algorithm filters the
extensions of formats and decides what kind of data should be prepared. For
example, it creates two-dimensional camera and different tools for pictures and
for triangulated geometry. Furthermore, it supports point clouds formats saved
in PLY and XYZ formats, with complete color information. Attaching six JPEG
files with proper names (directions like bottom, right etc.) results in creation
of cubic view with panoramic view, based on those files.
Window with 2D & 3D data consists of GUI controller and main view with
the loaded source. When having more than one geometry object in the same
Activity (at different level of detail), it loads the newest one, but it is possible
to change displayed model from drop-down menu. Additionally, the user can
control the settings of lightning, show geometry faces and vertices or change
material’s properties in real time.
Moreover, we designed a technique of describing those 3D objects and 2D
sources in WebGL canvas in three-dimensional space. It is possible to create
(a triangle, quad or any shape) or choose (from existing faces) an area that
should be featured. It is helpful for modellers or historians to highlight any
interesting part of the model. Then, they can annotate it because of possible
doubts, questions or additional data to be modelled, which can be connected
with events in database.
Fig. 4. Activity event during fireplace’s annotation.
5 Guidelines for 3D Modelling and Texturing

There are many different software products for 3D modelling, what causes a
lot of problems to be solved. First of all, there are many ways of calculating
and saving geometry’s properties like normals or faces. What is more, during
exporting/importing, the pivots of the scene are positioned differently due to
changed Y and Z axes. Some of the 3D software are allowed to store negative
texture’s coordinates, what is not appropriate for Web technologies.
We prepared the requirements for 3D modelling with an emphasis on Internet
applications.
– Due to backward hardware compatibility (16-bit vertex buffer), it is recom-
mended to not exceed the number of 65,536 vertices per model’s segment.
Splitting large models (e.g. >2 M vertices) into few smaller segments (eg. file
that contains of palace’s faade and roof could be splitted into north & south
parts of faade and roof separately) is worth considering. It makes much faster
and more convenient to store every object separately, because of describing
and displaying in real time.
– It is needed to export object with normals and texture coordinates. Modeller
should render an item to check if normals and textures are prepared well.
Fig. 5. The influence of texture’s dimensions on textured object.
– Triangulated mesh or polygons are necessary to display an object properly.

Interchargable formats (as OBJ) are not allowed to use quads or parametric
meshes.
– WebGL supports squared textures with dimensions of power of 2 (n × n,
where n2 ), see: Fig. 5.
– There is also prepared nomenclature for unique objects’s names. It stores the
data about main topic and what the part it is (e.g. Friedrichstein Manor-
House Roof − > F MANH ROOF). The directory and texture’s names are
also given: NameOfModel Date Version Textures/NameOfModel Date Ver-
sion TypeOfMaterial e.g.: F MANH ROOF 16-04-04 Ver2 Textures/F
MANH ROOF 16-04-04 Ver2 Tiles.jpg
– While using SketchUP, it usually gives improper results after exporting to
OBJ file. We found out that COLLADA is more accurate for SketchUP based
geometry, neverless it is much more complex to display and retrieve such
geometry.
Fig. 6. Prototype of the Virtual Museum: integration of interactive semantic 3D objects

in WebGL.
Fig. 7. Prototype of the Virtual Museum. Foreground: interactive, procedural real time
generated trees; background: billboarding trees.
6 Virtual Museum
The Virtual Museum (VM) connects many techniques and standards that are
commonly used in virtual reconstructions. There are existing and ready-made
solutions like X3D, which allow to create a virtual world. Regardless, the pos-
sibilities of such one are not enough for our researches and limited by external
Fig. 8. Rendered realistic scene with 3D models of Schlodien Palace (LumenRT).
libraries. However, there are no attempts to create a container for 3D geometry

that would compress needed files, reduce the time of decoding such data or com-
plexity of existing models. In fact, our solution also uses a lot of authors’ ideas
that were prepared for “virtual reconstructions” and solve the problems men-
tioned before. This approach exemplifies the adoption of new web standards and
gaming technologies, such as WebGL or Stage 3D, that extend the possibilities
of visualisation of 3D data, among others.
VM’s interactive form allows to visit non-existing places like Garden Room
in the destroyed manor house Schlodien (Fig. 6). The real-time displayed map
(1) with surroundings, supports the navigation and orientation on the user side.
The whole scene is composed of actual 3D geometry (saved in PNG file by
author’s algorithm) with textures. After picking some interesting (for user) ele-
ment, e.g. Fireplace, the semantically enriched content is loaded from the graph
database. As Fig. 6 point (2) shows there is a visual information (images, movies
used for the 3D reconstruction by the modeller) connected with this specific ele-
ment. Moreover, there is a paradata (the data describing the intellectual process
during the interpretation and computer-based reconstruction), medadata and
textual content (3) with complex data about historical background, names in
different languages and a link that directs to database’s page of this model
within the VRE www.patrimonium.net. Some of the places are associated with
a panoramic view (4) of existing state what makes it possible to compare past
and digital 3D representation with present outlook. There is also a list of initially
prepared “Point of Interest” which can be changed/generated by the user (add
or remove) for future use. In fact, if the user wants to share some view with
others or come back to the same place with same view, the website link contains
a position and look at vector, which is called Deep Linking.
The technique of splitting models with a huge number of vertices is frequently

used to reduce the file size; then a portalling method [8] can be applied. There
is also a LoD method [11], which requires few representations of the same object
with different number of vertices. If the user is far away from the model, algo-
rithm loads the object with the smallest number of vertices. When this distance
is decreasing, a model with more vertices should be loaded.
We have tested the LoD method in combination with a matrix of Level
of Information (in reference to the detail of sources) and Level of Hypothesis.
However a promising implementation still needs to be improved [7].
Prepared models for real-time calculated scenes can be also used for high res-
olution rendered images and movies (Fig. 8). For this purpose we use LumenRT
software, which allows to create more realistic environment with authentic
objects that has been measured and examined by art historians, architects and
3D modellers. There has been used more advanced light and shadow effects, more
complex vegetation and living animals. It is not possible to reach such results in
real-time with modern Web technologies on an ordinary computer. Previously
prepared and rendered images or movies can be easily used during exploring
WebGL environment, which appears to be a golden mean between quality and
efficiency.
7 3D Modelling of the Park and the Landscape
According to well known WebGL restrictions and hardware compatibility, we also

decided to generate trees in real time. It has been decided to use Open Source
library called proctree.js which generates complicated trees using GLGE - library
for direct access to OpenGL. Every tree’s type is described by 25 parameters
(including height, size of leaves, root, drop amount, twig scale, textures etc.)
and its position. An average number of vertices per one object is about 4 000,
depending on it’s parameters.
The whole prototype of Schlodien VM consists of about 350 trees what leads
to great number of vertices in this case. It is not possible to display that com-
plicated geometry in real time, so we used conjunction of two well known tech-
niques: billboarding (two crossed textures) and LoD (Level of Detail). Basic
version of LoD technique needs to load whole geometry at the beginning and
hide or display necessary ones at demand. It also caused displaying problems on
average laptop and stopped working while displaying about 60 trees. This lead
to introduce modified LoD technique that we called Progressive LoD (PLoD)
which bases on billboardings that are swapped by more complicated, real time
generated objects while going closer to them (Fig. 7). It appeared to be very
efficient method, allowing great interactive and seems to be golden meaning
between performance and appearance.
The terrain placed in Schlodien VM is based on actual cartographic map,
prepared by specialists. We have prepared high poly representation of the map
with usage of height map technique. It built high resolution model with real
dimensions and curvature of space according to map’s colors. Then, there has
been created few geometries at different levels of detail. It has been decided
to use the terrain model at low level of complexity due to restrictions of Web
technologies.
8 Conclusions for Combined Schema and Prospects on

Common Application Ontology
The increased digitisation efforts within the CH sector (e.g. in museums,

archives, libraries, research institutions, etc.) request strategies for extended
documentation standards serving the transparency (validity), interoperability,
long-term access and preservation of the information, in particular of the 3D
data sets.
Beside domain-driven documentation standards in museums, archives and
libraries, applied documentation standards for “digital 3D preservation”, spe-
cialised to capture and express a broad range of requirements for 3D digitisation,
already exists. At present we recognise the lack of documentation standards for
“digital 3D reconstruction”, based on a different procedure in data acquisition
and data processing, and leading to various 3D visualisation application.
The interactive visualisation of integrated 3D point clouds and hand-
modelled 3D data sets accompanied by the “semantic 3D annotations” tool
(under development) will have a significant impact on immediate research on 3D
data sets, the documentation and dissemination of our common “Digital Her-
itage” as the representation of the CH in the broader means. The mid-term chal-
lenge is to improve the documentation of the provenace and paradata, recently
expressed in the WissKI “field-based entry” and the “free text input”.
Our proposed solutions for the interactive integration of semantic 3D objects
are under development, but reached results promise to solve current problems
with transferring 3D data sets in real time. There are many reasons to continue
researches in this way, but Internet technologies (like WebGL) are demanding,
needing many techniques of optimization to reach the main aim.
Acknowledgments. We would like to thank our project partners, in particular Oliver

Hauck, Martin Scholz and Mark Fichtner for the cooperation in the design of the data
model and the VRE.
The international, collaborative research project and the resulting Virtual Research
Environments (patrimonium.net) under discussion, as well as the findings concerned
with the prototype of the Virtual Museum, are funded by the German Leibniz Associ-
ation from 2013 to 2016.
References
1. 3D-Icons: 3D-icons guidelines (2015). http://3dicons-project.eu/eng/
Guidelines-Case-Studies/Guidelines2
2. COLLADA: Documentation (2016). http://www.khronos.org/collada
3. Denard, H.: A new introduction to the London Charter. In: Bentkowska-Kafel,

A., Baker, D., Denard, H. (eds.) Paradata and Transparency in Virtual Heritage,
Ashgate, pp. 57–71 (2012)
4. Favro, D.: Se non vero, ben trovato (If not true, it is well conceived): digital immer-
sive reconstructions of historical environments. J. Soc. Arch. Hist. 1, 273–277.
University of California Press 71.3 (2012). Special Issue on Architectural Repre-
sentations
5. Görz, G., Schiemann, B., Oischinger, M.: An implementation of the CIDOC con-
ceptual reference model (4.2.4) in OWL-DL (2015). http://erlangen-crm.org/docs/
crm owl cidoc2008.pdf
6. Hauck, O., Kuroczyński, P.: Cultural heritage markup language - designing a
domain ontology for digital reconstructions. In: Proceedings of the 2nd Inter-
national Conference on Virtual Archaeology, Hermitage Museum, St. Peters-
burg/Russia, pp. 250–255 (2015). http://www.virtualarchaeology.ru/pdf/281 va
book2015.pdf
7. Kuroczyński, P.: Virtual museum of destroyed cultural heritage 3D documenta-
tion, reconstruction, visualization in the semantic web. In: Proceedings of the
2nd International Conference on Virtual Archeology, The State Hermitage, St.
Petersburg/Russia, pp. 54–61 (2015). http://www.virtualarchaeology.ru/pdf/281
va book2015.pdf
8. Lowe, N., Datta, A.: A technique for rendering complex portals. IEEE Trans. Vis.
Comput. Graph. 11(2005), 81–90 (2005)
9. Patrimonium: Virtual reconstructions (2015). http://www.herder-institut.de/go/
dp-962a8d
10. Scholz, M., Goerz, G.: WissKI: a virtual research environment for cultural heritage.
In: De Raedt, L., et al. (eds.) 20th European Conference on Artificial Intelligence,
ECAI 2012, Proceedings. IOS Press, Amsterdam (2012). http://wisski.eu/, http://
www.dh.cs.fau.de/IMMD8/staff/Goerz/ecai2012.pdf
11. Tan, K., Daut, D.: A review on level of detail. In: Computer Graphics, Imaging
and Visualization, pp. 70–75 (2004)
12. Trevett, N.: 3D transmission format. In: NVIDIA (2013)
13. TUD: Technical University of Darmstadt, Digitale Rekonstruktionen (2016).
http://www.dg.architektur.tu-darmstadt.de/dg/forschung dg/digitale rekonstruk
tionen/projekte.de.jsp
14. WebGL: Documentation (2016). http://www.chromeexperiments.com/webgl
Full Paper: The New Era of Museums
and Exhibitions: Digital Engagement
and Dissemination
CultureCam: An Interactive Search Tool
for Small Image Galleries
Sergiu Gordea1(B) , Michela Vignoli1 , and Sanna Marttila2

1
AIT Austrian Institute of Technology, 1220 Vienna, Austria
{sergiu.gordea,michela.vignoli}@ait.ac.at
2
Aalto University, Helsinki, Finland
[email protected]
Abstract. As a result of digitization initiatives in recent years, most gal-

leries hold digital copies of their masterpieces. In order to attract more
visitors, public galleries are interested in advertising their content on
websites and tourist-centric applications deployed in public spaces. The
online version of CultureCam has the goal of stimulating the reuse of cul-
tural heritage content by creative designers. In this paper, we present the
Interactive Installation version of CultureCam tool, which has the goal
of attracting the interest of public users when exploring public galleries.
It concentrates on enhancing the user experience, by offering access to
the images in an immersive environment, using an intuitive, easy-to-use
tool that supports touch free interaction with the gallery content. A
novel image similarity search algorithm was developed in order to adapt
to user expectations when searching in small image datasets. The user
feedback collected from exhibitions in different European cities indicates
a very high acceptance of the CultureCam tool by the public. The intu-
itive and seamless interaction with the tool, as well as the automation
and enhancement of the search algorithm are the main improvements
over the previous version of CultureCam.
Keywords: CultureCam · Interactive exploration · User experience ·

Similarity search · Cultural heritage
1 Introduction
CultureCam is an image similarity search tool, developed within the context of

the Europeana Creative project1 . The prototype has been designed to support
creative designers, and to increase the re-use of visual content that has a great
potential to inspire creation of new artifacts. The online version of the Culture-
Cam tool2 makes Europeana public domain image content easily accessible for
the open craft and media designer communities. By employing a search using
query by example approach we aim to overcome the limitations of free text search
1
http://pro.europeana.eu/web/europeana-creative.
2
http://culturecam.eu.

DOI: 10.1007/978-3-319-48496-9 25
310 S. Gordea et al.
(e.g. lack of descriptive metadata, semantic miss-match, complex search queries)

and offer an effective solution for visual content retrieval.
The visual artworks, which compose the dataset and are made accessible
online through image search, were manually curated by professional designers
according to their quality and reusability criteria. The relatively small dataset
used within the tool includes 3500+ artifacts selected from bio-diversity collec-
tions, illustrations from old books, portrait and landscape paintings, as well as
various design and poster artifacts. The user evaluation presented in [vignoli2015]
indicates that the tool was found to be quite useful and appreciated by the pro-
fessional designers for exploring reusable and inspiring image content. However,
in order to improve the effectiveness of the proposed approach for a larger public,
the search scenario needs to be adapted to the requirements and expectations of
regular users [gordea2014]. Additionally, the input used by the search algorithm
needs to be acquired in a more controlled environment, in order to reduce the
negative influence of background colors and the variation of light exposure on
the search precision.
The Interactive Installation version of the CultureCam tool presented in this
paper aims at enhancing the user experience in a scenario that concentrates on
interactive exploration of image content available in small visual art galleries.
In comparison to the online version of the tool, the new prototype introduces
important enhancements by providing:
– An intuitive and engaging interaction using a Kinect device. The users are
invited to explore the gallery through a seamless search and to play with the
content using a touch free user-system interaction;
– An immersive user experience using high quality projections on large surfaces
in a dark room;
– An efficient color based search algorithm using HSV image Histograms.
The concrete solutions proposed for addressing these challenges are described
in the main part of the paper, which is organized as follows: Sect. 2 presents the
related work, describing concepts and artifacts that are reused in CultureCam.
The detailed description of the proposed approach, the CultureCam prototype
and the concrete solutions contributed by this paper are presented Sect. 3. The
image search algorithm specially designed to comply with the requirements of
the Interactive Installation version of CultureCam is presented in Sect. 4, while
preliminary user feedback on the tool is discussed in Sect. 5.2. The conclusions
of this paper and the future work are summarized in Sect. 6.
2 Related Work
Given the contributions of the current paper, the related work is structured
in three categories, presenting similar approaches employed for: user experi-
ence improvement, advanced image visualization techniques, and similar image
retrieval algorithms.
CultureCam: An Interactive Search Tool for Small Image Galleries 311
User experience. Van der Sluis, van den Broek and van Dijk state that an
Information Retrieval system should be human-centered, as it solves the infor-
mation need of its user, with its user [sluis2010]. They suggest a Human-Centered
Model of Information Retrieval eXperience (IRX), which incorporates different
features into a coherent relevance model. They approach IR from a User eXpe-
rience framework (UX) perspective with a special focus on the emotional fac-
tors. UX is a fuzzy concept and is often defined as technology use beyond its
instrumental value [sluis2010]. Hassenzahl and Tractinsky divide UX in three
factors: Aesthetic and hedonic factors, emotional factors, and experiential fac-
tors [hassenzahl2015]. All these factors should be taken into consideration when
designing an IR system. Van der Sluis, van den Broek and van Dijk stress that
users develop negative emotions if they experience difficulties with the search
task. Thus IR systems should be user-friendly and meet their expectations in
order to be successful. To meet this requirement, CultureCam was build after
organizing a so-called co-creation workshop in which the requirements and needs
of professional designers were collected. Furthermore the selection of the image
dataset was guided by the professional designers.
Visualization. Many visualization approaches create static compositions of var-
ious images. For instance, Manovich analyzed and visualized over 50.000 Insta-
gram photographs shared in Tokyo over a few consecutive days [manovich2014].
The progression of activities by the people during the day is reflected by varying
colours and brightness of the images. The visualization creates an “aggregated
documentary” constructed from thousands of activities documented on Insta-
gram.
Other projects like RockViz offer a more interactive content visualization.
The project combined data of about 1994 significant Rock albums from Rovi and
visual analysis data from the gathered album covers. The images were organised
in various experimental web-based image plots. The user can filter the images
by group, title, or year, and click on the images for a larger view and details
[garcia2014].
In collaboration with Gravity Lab, the Software Studies Initiative developed
an interactive media visualization software running on the visual super computer
HIPerSpace [yamaoka2011]. The software can load up to 10.000 images and
interactively create a variety of media visualizations in real time. The images
can be analyzed and displayed in (multiple) image plots or scatterplots. This
technology allows analysis of single or multiple datasets and has been successfully
applied to large collections with hundreds of data sets, e.g. a collection of 883
manga titles with more than 1 million images [yamaoka2011].
Image retrieval. The diversity and heterogeneity of the Europeana collections
as well as poor textual descriptions of individual collections make a targeted text-
based search challenging. In terms of image items the content and particularities
of the represented objects are poorly described, which limits the possible text-
based search output considerably [gordea2014].
Content-based retrieval services such as the CultureCam provide comple-
mentary solutions to overcome the limits of text-based search. The tool follows
a search by example approach and includes a tactile centric interface for interac-
tively browsing a curated dataset with pre-selected images from the Europeana
image collection. The dataset is composed of items available in higher resolu-
tion and without copyright restrictions, as this would otherwise prevent their
re-use. Also, the images contain rare patterns as well as beautiful shapes and
colours. With the online version of CultureCam the user can take a picture of an
object with the computer’s or a mobile device’s webcam, whereupon the appli-
cation searches the index for images with similar colours, shapes and patterns
[vignoli2015].
In the tool, a nearest neighborhood algorithm is used to reduce the search
space and improve the execution performance at runtime. This is achieved by
selecting a pivot set and computing the distances between each indexed image
and the pivots. At runtime, the search results are ordered by their similarity
relative to the pivot set. A detailed description of the feature extraction and
indexing process is available in [amato2011]. The new search algorithm intro-
duced in Sect. 4 follows the same indexing approach, but it uses a clustering
approach based on the centroids of non-uniform color beans in HSV space. 54
Color clusters and 4 grayscale clusters were empirically defined basing on the
visual representation of the 6 main color sections of HSV space.
3 Interactive Installation Approach
As explained above, CultureCam offers a more straightforward and intuitive way

of exploring image content than the more traditional text based search on the
Europeana portal. It stimulates and offers new possibilities for creative re-use
of Europeana content, and facilitates the creation of new content visualizations
with shapes and patterns, which have only a small probability of arising in other
frameworks [dorin2009]. The followup Interactive Installation version of Culture-
Cam brings the content closer to public user and public spaces and extends the
image content exploration to a physical, touch free, and immersive user experi-
ence.
The objectives of the prototype were to create an experience-based (i.e.
through stimulation of emotions like fun, fascination, intrigue) version of Cul-
tureCam tool tailored for public or private exhibition spaces. It is designed as
a stand-alone system, which can be easily adapted to specific collections and
venues. The prototype presented here uses the same image dataset as the online
version. However, the tool was designed to easily change the image content and
create exhibitions of existing digital collections of galleries or museums.
3.1 An Immersive User Experience
The overview of the physical system is presented in Fig. 1. A high definition

projector is used to present the image content on large surfaces in order to place
the user in an immersive environment. When projecting on a surface like large
walls in public institutions, the projector is typically placed behind the user at
a height that eliminates the obstruction of the projection by the human body.
The exploration of the image content is triggered by the automatic capture of
pictures with the camera of the Kinect device3 . The movement sensors are used
to detect when the user enters a square which is marked on the floor as being
the focus position of the camera. The image search service and the interactive
image rendering application, run on a commodity computer that is connected
with the projector and the Kinect device.
Fig. 1. Culture Cam: physical system
In order to reduce the influence of background colors, a black molton fabric

is placed behind the user, and the black color is weighted to narrow scores by
the image search algorithm. As seen in the sketch, the Kinect device is in front
of the user, at a distance that takes pictures of the upper part of the user’s body.
Consequently the captured images will represent the pullovers and shirts worn
by the human user. However, different artifacts like scarfs or images printed on
paper can be used by users to explore different subsets of the image collections.
The most impressive user experience is obtained in dark rooms, simulating the
cinema environment (see Fig. 2). In this case a source of directed light needs to
be placed near the Kinect to ensure the required light exposure for capturing
decent quality pictures.
The Kinect device is not only used for capturing images of one person at
a time. The Interactive Installation also features gesture-based interaction for
other visitors who want to interact with the image search results displayed on the
wall. Specifically movements of the hands and feet are captured by the Kinect
device. By pointing to a certain area of the wall, the images in that area are
enlarged and animated in a three-dimensional fashion giving the user the feeling
of touching or getting closer to the images. As a further visual effect all the
3
http://www.xbox.com/en-US/xbox-one/accessories/kinect-for-xbox-one.
Fig. 2. Gallery wall of projected Europeana images in National Gallery of Denmark.

Photo by Christina Holm CC BY-SA 4.0
images currently displayed on the wall rotate when the Kinect captures a new
image, which indicates the process of similarity search for a new set of images.
3.2 Touch-Free and Engaging Interaction
As explained above, the user can interact with the CultureCam via a Kinect
sensor covered by a black box, which scans/photographs the visitor. Once the
motion sensor of the Kinect has detected a person standing on the marked “scan-
ning spot”, a picture is taken and analyzed in order to retrieve Europeana images
matching the colors and patterns in the photo. In a larger area around the “scan-
ning spot” up to six people can interact with the projected images by making
movements to enlarge and animate the images. All in all the installation gives
the users a highly dynamic and spatial experience (see Fig. 3).
For the exhibition at the National Gallery of Denmark 3,231 public domain
and CC0 licensed images from the Europeana collection were used. Due to the
large size of the wall, the relatively small data set, and the impact of the sim-
ilarity functionality the images were repeated to fill out the projection area on
the wall as well as to provide the audience with a clear and immediate visual
understanding of the colour similarity concept of CultureCam. The interface of
the installation resembles the online version of CultureCam: the captured photo
of the person is displayed in a circular shape area in the center of the screen.
A more comprehensive report on the CultureCam product prototypes and
their evaluation can be found in the Europeana Creative deliverable D4.6 Deliv-
ery of the Design Pilot [marttila2015].
Fig. 3. Museum visitors interacting with the CultureCam installation using props
4 Efficient Search Algorithm Using HSV Color

Histograms
The novel search algorithm designed to serve the Interactive Installation is based
on HSV Color Histogram. The HSV representation of the color space is more
intuitive and easier to be sampled in color bins that have a clear functional
meaning for human user. However, as with most color representation models,
HSV is not a uniform color space4 . For computation of the color histogram the
color space was divided in 54 color bins and 4 grayscale bins.
4.1 Image Descriptor

The grayscale bins contains all colors with the saturation (S) and/or value (V)
smaller that 32, with the meaning of grouping values close to white, black, light
gray and dark gray colors. The color bins are built using the 6 base colors of the
HSV space: red, yellow, green, cyan, blue, magenta. Given that the lower values
for H and V are distributed over the grayscale bins, the rest of the values are
grouped in 3 bins for each. Consequently, the resulting bins include 54 sets of
colors (i.e. 6H * 3S * 3V).
The HSV Color Histogram of an image is computed by counting the number
of pixels that belong to each color bin. Given that the input images and the bin
are not of equal size, the image descriptor used for indexing and search purposes
is using the normalized HSV histogram (N H(i)), see Eq. 1. The normalization
of the histogram values (normScore(bi )) is computed by multiplying the ratio
of pixels from each bin (i.e. computed as percentage) in the picture with the
entropy of each bin. The bin entropy is the inverse ratio between the size of
individual bins to the average bin size, see Eq. 2.
4
https://en.wikipedia.org/wiki/HSL and HSV#HSV.
N H(i) = {normScore(bi ))|bi ∈ B, I} (1)

count(pi ∈ bi ) ∗ avgSize(b) ∗ 100
normScore(bi ) = , (2)
size(bi ) ∗ size(i)
Where bi represents a color bin of image I, with regard to the bin definitions B.
normScore(bi ) represents the normalized score for bin bi to which the pixels pi
are assigned. avgSize(b) is the average size of all b bin definitions computed as
number or pixels, size(bi ) counts the number of pixels pi associated to bin bi ,
and size(i) represents the size of the image i computed in pixels.
4.2 Indexing and Search Algorithm

For building the image search index, we adopt a similar approach as Amato
et al. [amato2011]. We aim at taking the benefits of the advanced technologies
developed for supporting high-performance and scalable search services for text
documents, for which Lucene library is the most frequently used implementation.
The normalized color histogram descriptor is represented as a map, in which
the hexa representations of the color bin centroids are used as keys and the
normalized scores represent the values. This is an equivalent representation to
the bag of words used for text indexing, where the integer value of the normalized
score serves as term frequency. Consequently, the descriptors or all images in the
dataset are used to build the image index. In the same way, the descriptors or
query images are converted to text search queries, in which the normalized scores
are used for boosting terms in the search queries. At runtime the similarities
between the query image and the dataset items is computed using the cosine
similarity as shown in Eq. 3.

similarity(q, x) = (tf (c ∈ hist(x)) ∗ boost(c ∈ hist(q)) ∗ idf (c)2 ), (3)
c∈hist(q)
Here the similarity(q, x) represents the similarity between the query image q
and the indexed image x. c represents the identifiers of color bin centroids in the
normalized color histogram of query image q (i.e. hist(q)). The tf (c ∈ hist(x))
represents the term frequency of color bin centroid c in the color histogram of
image x, and it is computed using the normalized score of the color bin (i.e.
see Eq. 2). Also the boost of the centroid c in the color histogram of the query
image q is computed using the normalized score for the given bin. While idf (c)
represents the inverse (document) frequency of centroid c and it is an index
constant. This constant has a marginal influence on the ranking of the search
results, therefore, implementations may set this to 1, which is equivalent to a
uniform distribution of centroids over the indexed documents.
5 User Experience Evaluation

5.1 User Experience for Online Version
The user experience testing for the online version of CultureCam was carried out
in a controlled environment using professionals from the creative design domain.
Fig. 4. TagCloud: Criticism
Fig. 5. TagCloud: Enhancement suggestions
The goal was to verify to which extent the tool satisfies the expectations and
needs of users and what kind of new functionality the users would recommend
for subsequent versions. The user feedback was collected in a two step process
including:
– An online survey collecting structured information regarding the usability and

efficiency of the tool
– User interviews collecting explicit feedback regarding the user experience and
ideas for further improvements
There were 10 users who took part in the evaluation, 4 of them in

Barcelona/Spain and 6 in Helsinki/Finland. They were active professionals hav-
ing the following roles in their daily work: graphic designer, illustrator, visual
artist, art director, editorial designer, architect. The users were provided with
objects regularly used by designers and they were also encouraged to use their
own artifacts for evaluating the value of CultureCam with regard to their daily
work. The results of the online survey show that the great majority of the users
had a positive user experience. The negative feedback was related to the navi-
gation and the limited interaction with the image content (for more details on
the evaluation see evaluation report [senabre2015]).
The narrative feedback collected was divided in tree categories, positive
feedback, criticism and enhancement suggestions. The TagCloud representation
including the top 20 words for the criticisms and enhancement suggestions are
visually presented in Figs. 4 and 5 respectively. This type of feedback was par-
ticularly important for development of the follow-up version of the tool, which
is the Interactive Installation. The criticism of the interaction with the tool was
mainly related to the handling of the camera, the low resolution of the thumb-
nails and the lengthy way of accessing higher quality images. There were also
criticisms regarding the layout, the tooltip texts and the navigation icons. Sug-
gestions for further enhancement of the Culture cam tool were collected and
they include requests for functionality like:
– Storing navigation history and enhancing the sharing functionality (e.g. adding
possibility to share with pinterest, sharing images, more share buttons)
– Requirements for improving the relevance of search results
– Support for more interaction with search results, like and grouping images in
mood boards
The testers gave clear affirmative answers to the questions related to the
product’s new way of approaching and re-using digital heritage. In particular,
the testers confirmed that the platform makes re-usable high quality content for
designers available and accessible, and that the product compiles already existing
content in a uniquely and easily navigable platform, encouraging creative re-use.
Two users compared it to Shazam5 , a music exploration tool, but for images.
One user indicated that CultureCam offers many opportunities for museums,
and another user liked the idea behind the tool, but he was puzzled that the
results shown by the tool when capturing images of the same scene (with the
same or with different devices) are not always the same ones.
5.2 User Experience Observations for Interactive Installation

Version
The Interactive Installation version of the CultureCam was implemented to

address the weaknesses of the online version, to enhance the user experience,
and to bring it closer to public users. This version was not evaluated in a con-
trolled environment, however it was presented at various events. In May 2015
the interactive CultureCam was installed at the National Gallery of Denmark
at the Set art free! Open Culture event. The visitors of the exhibition could
interact with the tool by standing in front of a black box covering a Kinect
motion sensor, which scans/photographs the visitor. The installation projected
CultureCam’s search results based on the interacting person’s clothes on a large
wall in the gallery. Because of the enormous projection, which filled the whole
wall, the individual experience of the visitors triggering the search turned into
a shared experience by the audience [marttila2015].
The CultureCam installation was well attended during the Set art free! event.
People of all ages were attracted due to the high degree of playfulness of the
installation. Two large posters introducing the CultureCam prototype as well
as the Europeana Creative project in Danish and English were displayed. The
interaction and subsequent response to human movements on the wall were intu-
itively understood by the audience. Some assistance was requested to understand
exactly why and when certain positions/movements trigger specific reactions by
the tool. In particular questions related to the exact positions in regards of the
Kinect to get the best capture result were posed by the visitors.
The installation stimulated visitors’ emotional responses and added a unique
aesthetic value to exhibition space and experience. The visitors seemed amazed
about and overwhelmed by the transformation of the gigantic wall of e.g. entirely
5
http://www.shazam.com.
Fig. 6. Selected online user feedback
blue images to entirely red ones. For this event the a new search algorithm was
developed which use only color descriptors, as the shape similarities are not well
enough described by global image descriptors. This improved the intuitiveness
of the installation dramatically leading to the result mentioned above. Another
factor was the influence on the projections by the decreasing natural light in the
room, which led to an aesthetical transition of the installation during the evening.
The aesthetic impact played a significant role in the total user experience. The
CultureCam installation generated very positive user feedback in social media,
in which the users characterized the installation as cool, stunning, and popular
(see Fig. 6).

This paper presents the CultureCam tool that makes visual navigation through
selected image content in Europeana as simple and intuitive as possible. The
current version of CultureCam uses a collection of 3700 items that were curated
to be freely reusable and inspiring for creative designers, as well as to stimu-
late the interest of public users of digital cultural heritage content. Within this
paper we focus on the user experience improvements, which were implemented
in the Interactive Installation version of the tool. Following the user feedback
collected on the previous version of the tool, different techniques were integrated
to enhance user experience, including: presentation of cultural heritage content
in an immersive environment, touch free interaction, and a specialized image
search algorithm. The tool was exhibited at a couple of events in different Euro-
pean cities, where the new version of the tool was very appreciated. Besides
winning the audience award at the Culture Jam 2015 in Vienna, CultureCam
was categorized as an outstanding product in social media and various blog
posts6 ).
Future work will concentrate of further development of the search algorithms
to enhance the precisions of the search, by integrating the usage of MPEG local
descriptors (i.e. ISO/IEC CD 15938-13) and deep learning models for classifica-
tion, clustering and indexing. Additionally, we aim at enhancing the impact of
the tool in the cultural heritage domain, by deploying the Interactive Installation
in different Galleries, using their own datasets. For the online version of the tool
the ambition is to integrate it within the portal of the Art-History Collection.
Acknowledgments. This work was partially funded by the EuropeanaCreative

project, co-funded by the Commission of the European Communities under the
ICT Policy Support Programme (ICT PSP). This publication reflects only the
author’s/authors’ views and the European Union is not liable for any use that might
be made of information contained therein.
References
[gordea2014] Gordea, S.: An image similarity search for the European digital
library and beyond. In: TPDL 2013 Selected Workshops, Communi-
cations in Computer and Information Science, vol. 416, pp. 190–196
(2014)
[amato2011] Amato, G., Bolettieri, P., Falchi, F., Gennaro, C., Rabitti, F.: Com-
bining local and global visual feature similarity using a text search
engine. In: Content-Based Multimedia Indexing (CBMI), pp. 49–54
(2011)
[dorin2009] Dorin, A., Korb, K.: Improbable creativity. In: Dagstuhl Seminar
Proceedings of Computational Creativity: An Interdisciplinary App-
roach, nr. 09291 (2009)
[vignoli2015] Gordea, S., Vignoli, M.: Culturecam: visual exploration of cultural
heritage content by professional designers. In: Multimedia Expo
Workshops (ICMEW), pp. 1–6 (2015)
[hassenzahl2015] Hassenzahl, M., Tractinsky, N.: User experience-a research agenda.
Behav. Inf. Technol. 25(2), 91–97 (2015)
[manovich2014] Manovich, L.: Watching the world. In: Aperture, nr. 214 (2014)
[marttila2015] Marttila, S., Holm, C., Hyypp, K., Gordea, S., Laakso, N., Stokes, L.,
Senabre-Hidalgo, E.: Delivery of the design pilot. Technical report
(2015). http://pro.europeana.eu/files/Europeana Professional/
Projects/Project list/Europeana Creative/Deliverables/eCreative
D4.6 NISV v1.0.pdf
[garcia2014] Reyes-Garcia, E.: Explorations in media visualization. In: Extended
Proceedings of the 25th ACM Conference on Hypertext and Hyper-
media Hypertext. ACM Press, New York (2014)
6
https://twitter.com/hashtag/culturecam.
[senabre2015] Senabre, E., Utratel, A.: Pilot and infrastructure evaluation

report. Technical report (2015). http://pro.europeana.eu/files/
Europeana Professional/Projects/Project list/Europeana Creative/
Deliverables/eCreative D6.3 platoniq v1.0.pdf
[sluis2010] van der Sluis, F., van den Broek, E., van Dijk, B.: Information
retrieval eXperience (IRX): towards a human-centered personalized
model of relevance. In: Web Intelligence and Intelligent Agent Tech-
nology (WI-IAT), vol. 3, pp. 322–325 (2010)
[yamaoka2011] Yamaoka, S., Manovich, L., Douglass, J., Kuester, F.: Cultural ana-
lytics in large-scale visualization environments. Computer 44(12),
39–48 (2011)
Learning Algorithms for Digital Reconstruction
of Van Gogh’s Drawings
Yuan Zeng(B) , Jiexiong Tang, Jan C.A. van der Lubbe, and Marco Loog
Delft University of Technology, 2628 CD Delft, The Netherlands

{Y.Zeng,J.Tang-5,J.C.A.vanderLubbe,M.Loog}@tudelft.nl
Abstract. Many works of Van Gogh’s oeuvre, such as letters, drawings

and paintings, have been severely degraded due to light exposure. Dig-
ital reconstruction of faded color can help to envisage how the artist’s
work may have looked at the time of creation. In this paper, we study
the reconstruction of Vincent van Gogh’s drawings by means of learn-
ing schemes and on the basis of the available reproductions of these
drawings. In particular, we investigate the use of three machine learn-
ing algorithms, k-nearest neighbor (kNN) estimation, linear regression
(LR), and convolutional neural networks (CNN), for learning the recon-
struction of these faded drawings. Experimental results show that the
reconstruction performance of the kNN method is slightly better than
those of the CNN. The reconstruction performance of the LR is much
worse than those of the kNN and the CNN.
Keywords: Van Gogh’s drawing · Drawing reconstruction · Reproduc-

tion · Machine learning
1 Introduction
Van Gogh is famous for his canvas paintings, but the largest part of his oeuvre
consists of works of art on paper [1]. For this he used different instruments and
materials: pencil, crayon, ink, watercolor and oil paint, often in combination.
However, many drawings of Van Gogh have dramatically deteriorated over the
last century. This especially holds for the ink drawings, where degradation is
caused by ink fading due to light exposure among others. An dramatic example
can be found in Fig. 1, where the entire original drawing has been faded. In
addition to ink fading there are also other types of color changes. Many drawings,
which now have a brown color, were originally black, blue, or purple. In Fig. 1(b)
the original purple color is visible in the margin, since it was occluded by a frame
and therefore largely protected from the influence of light.
One of the goals of the interdisciplinary REVIGO-project (REassessing VIn-
cent van GOgh, funded by the Netherlands Organization for Scientific Research
NWO) is to predict the original and future appearances of art works by means of
advanced digital methods that integrate the results of both in-depth studies of
the colors used and their discoloration over time. This will help to envisage how

DOI: 10.1007/978-3-319-48496-9 26
Learning Algorithms for Digital Reconstruction of Van Gogh’s Drawings 323
the artist’s works of art may have looked at the time of creation. Furthermore, it
will help for the identification of appropriate conservation and restoration treat-
ment strategies as well as of implications for future preservation and display.
Although in this paper the focus is on Van Gogh, the problem of fading and
discoloration holds for most works of art on paper and manuscripts from the
19th century.
Fig. 1. (a) Reproduction of the present state of “Montmajour ”, made by Van Gogh in
1888; (b) The original aniline ink color in the margin of the present “Montmajour ”;
(c) Reproduction of “Montmajour ” which was made in 1928. (Color figure online)
This paper can be considered as an initial study of the possibility of digital

reconstruction based on reproductions of the original drawings in combination
with state-of-the-art supervised learning techniques.
In the course of the 20th century, many reproductions were made of Van
Gogh’s drawings, i.e., photographs, facsimiles, stencils, and so on. In some cases,
e.g., for the fountain (1889) and La Crau (1888), there are at least 10 to 20 high
quality reproductions dating from 1910 to 1928. Especially the early reproduc-
tions are important because they show the unfaded drawing. Most reproductions,
however, are black and white. Only from a limited selection of drawings and
watercolors colored facsimiles exist. Nevertheless, these reproductions contain a
large amount of information that could be exploited in reconstructing the orig-
inal works based on their present degraded appearance. The idea is that based
on various examples of degraded appearances and their corresponding original
(or at least less degraded) appearances, it may very well be possible to learn
the inverse process, at least partly. If so, also for unseen works, for which no
reproductions are known, reconstructions of their past appearance may be pos-
sible. In this paper, a first step is made to exploit the information content of
reproductions for the digital reconstruction of drawings.
Three machine learning algorithms are considered. As a simple baseline, lin-
ear regression (LR) [2] is considered. The more powerful, nonlinear methods con-
sidered are k-nearest neighbor regression (kNN) [3,4] and convolutional neural
networks (CNN) [5–7]. In this paper, next to the question to what extent recon-
struction might be possible at all, we are also interested in the performance of a
CNN in comparison with more traditional methods like LR and kNN.
324 Y. Zeng et al.
2 Data
For this initial study, a single faded present-day Van Gogh drawing together with
one reproduction has been chosen. The latter is a reproduction from the 20th
century for which the content and color are faded less and which is closer to the
original drawing. The drawing and the reproduction are obtained from the Van
Gogh Museum collection. In general, one drawing may have several reproduc-
tions, and those reproductions can be categorized based on two aspects: degrada-
tion level and reproduction method. Reproductions made with the same method
but different degradation level can show the degradation process over time. How-
ever, reproductions with different methods can severely affect the reconstruction
process, and thus affect estimation performance of the learning algorithms, since
color or even content of reproductions made with different methods can be differ
from each other. In this work we want to find out what is at all possible in a
simple prediction setting and therefore we study a basic setting with one input
(faded original) and one output (close to non-degraded original) image. Later
this work can be extended to one drawing and multiple reproductions.
For the data, we consider two images of the same drawing, where one is the
present faded version and the other is the reproduction. x and x are parts of
the faded image and the reproduction, respectively, which are used as training
images. y and y are corresponding parts in the faded image and the reproduc-
tion, respectively, which are used as test images. Then, given a test image y,
the object of learning algorithms is to learn the degradation process of x and
x, and thus predict y , where the degradation between y and y is the same as
the degradation between x and x.
3 Methods
In this section, we introduce framework for learning the degradation process

between x and x, and predicting target image y by using three different machine
learning algorithms, which are the linear regression (LR), the k-nearest neighbor
estimation (kNN) and the convolutional neural networks (CNN).
3.1 Linear Regression
Let us consider I × J dimensional training images x and x . Then, the learning

process of simple linear regression is given by
xij = a0 + a1 xij + ij , (1)
where i = {1, · · · , I} is the row index, j = {1, · · · , J} is the column index, xij is
the RGB value of the ith row and jth column pixel in x and ij is the estimation
error. Then parameters a0 and a1 can be estimated by minimizing the square
value of estimation error, that is
I
J
min 2ij . (2)
a0 ,a1
i=1 j=1
Later, given a test image y, a target image y can be estimated as ŷ = a0 + a1 y.
3.2 k-nearest Neighbor Estimation

The basic idea of the method in this subsection for estimating the target image
y is per-pixel k-nearest neighbor estimation. Assuming that a test image y

consists of R × C pixels. Then, for each pixel ŷrc with row index r = {1, · · · , R}
and column index c = {1, · · · , C}, we first search the k nearest pixels in x and

then predict the pixel ŷrc via simply averaging the k nearest pixels, i.e.,
k
1
ŷrc = x , (3)
k i=1 pi
where p = [p1 , · · · , pk ] is the index of the k closest pixels. The method for
searching the k nearest pixels in x for each pixel ŷrc
is based on measuring
the closeness between feature vector of the corresponding pixel yrc and feature
vectors of all pixels in x. The searching method can be described as follows:
1. For each pixel xij , based on s neighboring pixels of the pixel xij , we construct
a feature vector Fxij = [xi−s,j−s , · · · , xi,j , · · · , xi+s,j+s ]. In this case, we can
construct a s2 ×(I × J) dimensional feature matrix Fx of the training image x.
2. For the test pixel yrc , similar as the step 1), we construct a s2 × 1 dimensional
feature vector Fyrc based on the s neighboring pixels of yrc .
3. Measure the closeness between the feature vector Fyrc and the feature matrix
Fx via the standard Euclidean distance as
2
D = Fyrc I − Fx , (4)
where I is a 1×(I × J) dimensional vector with all ones, and D is a (I ×J)×1

dimensional vector.
4. Based on the index p= [p1 , · · · , pk ] of the k smallest values in D, select the
k pixels xp1 , · · · , xpk in the training image x as the k nearest pixels of the

estimated pixel ŷrc .
Here we mainly present the framework of the kNN for color reconstruction, while
the detailed theoretical basis and convergence properties of the kNN regression
is given in [3].
3.3 Convolutional Neural Networks

A typical single layer CNN consists of three basic parts: convolution, non-linear
mapping and pooling. It can be denoted as
l−1

alij = Pool ReLU Wl ∗ aij + bl , (5)
326 Y. Zeng et al.
where alij is the activation from lth feature maps with row and column indexes
l−1
(i, j). And if this is the input layer, the aij are the pixels of input images x.
l−1
And if it is output layer, aij are pixels of the predicted images x . Wl and bl are
learned weights and bias of convolution filters, respectively. ∗ is the convolution
operation. ReLU (x) = max (0, x) refers to the non-linear transform of a rectified
linear unit (ReLU). Pool (·) is the pooling operator for down-sampling the image
with certain strategy, e.g., Max Pooling and Average Pooling, which uses the max
or average value to replace the pixels of the same local region.
Based on the basic structure introduced above, in this paper, the specification
of the CNN layers for reconstruction of the drawings could be seen in Fig. 2.
Please note that, every convolution layer before pooling and the fully connected
layers are with ReLU layers on top, they are not explicitly shown here. Firstly, the
typical structure is stacked three times to detect different details from multiple
views. Specifically, the convolution is implemented twice before pooling. Then,
the fully connected layers are followed to further encode the high level features.
Since the convolution combined with pooling will gradually down-sample the
input, three deconvolution layers are build on top the fully connected layer, they
will reconstruct the image with its original resolution. The deconvolution is the
inverse transform of convolution, it inserts pixels with learned weight and bias
based on the value in each location. Finally, the square loss is used to measure
the distance between prediction from current drawing and its reproduction.
In general, the CNN treats the reconstruction as a pixel-level regression which
is performed by using three procedures: convolution, encoding and deconvolu-
tion. Proposed structure uses the current drawing images as input and directly
output the prediction. The function of overall degradation is expected to be
learned during the optimization rather than be formulated explicitly.
4 Experiments
In this section, we illustrate the estimation performance of the three machine
learning algorithms, which are the LR, the kNN and the CNN. First, we analyze
the trade-off between the estimation performance of the kNN method and the
three parameters (i.e., the amount of training data |x|, the amount of neighboring
pixels for constructing feature vector s and the number of nearest pixels k). After
that, the estimation performance comparison between the LR, the kNN and the
CNN is given.
To use the reproduction in Fig. 3(b) for reconstruction of the present state
drawing in Fig. 3(a), we first use control point registration method to align the
present drawing and its reproduction. This indicates that the registration per-
formance depends on the amount of and the accuracy of selected control points,
where more accurate control points leads to better registration performance.
Although the registration accuracy can be improved via selecting hundreds or
even thousands control points with enough accuracy, it is very difficult to obtain
perfect alignment of the drawing and its reproduction, since digital informa-
tion of the drawing and digital information of its reproduction are collected by
Fig. 2. Layer specification
using different devices. In this work, we focus on the study of reconstruction of

drawings via different machine learning algorithms, we thus simply register the
drawing in Fig. 3(a) and its reproduction in Fig. 3(b) with 30 manually selected
control points and use the as such aligned images for our study. In addition,
the dimension of the drawing in Fig. 3(a) and its reproduction in Fig. 3(b) is
6132 × 8176. To remove blank margin, we use effectively 5000 × 6500 pixels of
both the drawing and its reproduction in our experiments. The mean square-
error (MSE) is used as a measure to assess the estimation performance of the
learning algorithms. The MSE between the estimated image ŷ and the target
image y is given by
R C
1 2
MSE = |ŷ − yrc | . (6)
RC r=1 c=1 rc
In the following experiments, the 5000 × 6500 dimensional drawing and its
reproduction are divided into 50 × 50 sub-images, respectively. Three-quarters
of the 2500 sub-images are randomly selected as training images (in total 1875)
and the remaining sub-images are divided into validation set (70 sub-images)
and test set (555 sub-images).
4.1 Parameter Estimation
Since the estimation performance of the kNN method depends on the three
parameters: |x|, s and k, the trade-off between the estimation performance and
328 Y. Zeng et al.
Fig. 3. (a) Present version of Van Gogh’s drawing Old vineyard with peasant woman
with size 44.3 cm × 54 cm; (b) Reproduction of Old vineyard with peasant woman which
made in 1928 with size 41.9 cm × 51 cm.
those three parameters is analyzed in this subsection. Since the computational

complexity of the kNN depends on the dimensionality of feature vectors and
thus the amount of pixels in the training data, we construct training data by
randomly selecting n pixels per training sub-image (|x| = n × 1875), and the
test data in this experiment is constructed by randomly selecting 1 × 104 pixels
from the validation set.
Firstly, we evaluate the MSE as a function of the n and the s with fixed
k = 12. Figure 4(a) shows the simulation results in terms of MSE between the
estimated image ŷ and the target image y , where the MSE is decreased with
increasing n from 1 to 150, and the MSE is decreased with increasing the s from
1 to 9, but it is increased with s larger than 9. This indicates that the optimal
setup of the s is 9. Next, to evaluate the MSE as a function of the n and the
k, we set s = 9. The experimental results in Fig. 4(b) show that the MSE is
decreased with increasing k from 1 to 15, and it is decreased slowly when k is
larger than 10.
Fig. 4. (a) MSE versus the n and the s; (b) MSE versus the n and the k.
To further show how the three parameters affect the estimation performance
of the kNN, we randomly select one sub-image from the validation set as test
image y and its corresponding image from the reproduction as target image y .
We then evaluate the estimation performance of the kNN with eight different
Fig. 5. (a) Test image y; (b) target image y ; (c) ŷ with n = 5, s = 9 and k = 10,
and the MSE = 0.0061; (d) ŷ with n = 45, s = 9 and k = 10, and the MSE = 0.0060;
(e) ŷ with n = 145, s = 9 and k = 10, and the MSE = 0.0058; (f) ŷ with n = 145,
s = 9 and k = 3, and the MSE = 0.0084; (g) ŷ with n = 145, s = 3 and k = 10, and
the MSE = 0.0125; (h) ŷ with n = 45, s = 13 and k = 10, and the MSE = 0.0073.
setups of the n, the s and the k. The simulation results in Fig. 5 are consistent
with the analysis in Fig. 4, where MSE is decreased with increasing either the
n from 5 to 145, the s from 3 to 9 or the k from 3 to 10. In the following
experiments, we use n = 150, s = 9 and k = 10 for the kNN estimation.
4.2 Comparison Experiments
Here, we compare reconstruction process of the LR, the kNN and the CNN
in terms of MSE. In order to show visible estimation results, we first carefully
select 8 representative sub-images (e.g., images with different patterns, color and
texture) from the test set.
Figure 6 shows the comparison between the LR, the kNN and the CNN out-
puts of the 8 sub-images. Visually, we observe that in general the estimation
outputs of both the CNN and the kNN are much better than the LR method.
This is reasonable since the LR method uses a simple linear function to model the
relation between the outputs and the inputs, whereas the relation between the
drawing and its reproduction is not linear, see Fig. 7. Furthermore, the estima-
tion results for the sub-images (1), (2), (3) and (6) show that the CNN can learn
textural information and provide better visual outputs, while the kNN method
smoothly predicts output by averaging k nearest neighbors and thus loses tex-
tural details. The CNN also works much better than the kNN for estimating the
sub-image (8). This is because the kNN is based on local pixel-level optimization,
and thus the estimation performance of the kNN can be worse if pixels with close
RGB values in training data x relates to pixels with two or several different RGB
values in x . However, the estimation results of the CNN for the sub-image (3),
(4), (5) are worse than the kNN method. This can partly explained by the fact
that the contour of the local blobs in those sub-images could barely be seen due
to degradation, the CNN seems fail to detect the specific structure and shape,
i.e., the faded yellow half circles. From the perspective of the MSE in Table 1, it
is observed that both the kNN and the CNN predict better than the LR, and the
kNN performs better than the CNN for the sub-image (2), (3), (4), (5) and (6).
330 Y. Zeng et al.
Fig. 6. Estimation performance comparison between the LR, the kNN and the CNN.
Fig. 7. Red value of a sub-image in the drawing and its corresponding sub-image in
the reproduction.
Next, to compare general performance between the three learning algorithms,

we test all sub-images in test set and assess the estimation performance of the
three learning algorithms by averaging MSE (AMSE) of the M = 555 estima-
tions, that is
M
1
AMSE = MSEm , (7)
M m=1
where MSEm is the mean square-error between the mth estimated sub-image
and the mth target sub-image and can be computed via Eq. (6). To assess the
estimation performance for each color channel, we denote AMSER , AMSEG and
AMSEB as the AMSE of red, green and blue channel, respectively. The simula-
tion results are given in Table 2, where the estimation performance of the kNN
Table 1. Estimation performance comparison between the LR, the kNN and the CNN
in terms of MSE.
(1) (2) (3) (4) (5) (6) (7) (8)

MSELR 0.0039 0.0068 0.0105 0.0008 0.0011 0.0023 0.0377 0.0033
MSEkN N 0.0047 0.0042 0.0078 0.0004 0.0005 0.0036 0.0015 0.0024
MSECN N 0.0034 0.0065 0.0099 0.0005 0.0020 0.0044 0.0015 0.0018
Table 2. The AMSE of the LR, the kNN and the CNN.
LR kNN CNN
AMSER 0.0113 0.0067 0.0093
AMSEG 0.0113 0.0068 0.0094
AMSEB 0.0112 0.0068 0.0092
and the CNN is much better than the LR, and the kNN performs better than
the CNN.
5 Discussion and Conclusions
In this paper, we introduced three classic machine learning algorithms for recon-
structing faded color of Van Gogh’s drawings: the linear regression (LR), the
k-nearest neighbor estimation (kNN) and the convolutional neural networks
(CNN). We first analyzed the trade-off between the estimation performance of
the kNN as function of the three parameters (e.g., the amount of training data,
the dimensions of feature vector and the k). The experiments have shown that
better reconstruction performance is obtained by using the estimated optimal
parameters for the kNN. Furthermore, we compared the reconstruction perfor-
mance between the three learning algorithms. The experimental results have
shown that the reconstruction performance of the LR method is worse than the
kNN and the CNN, and that the kNN and the CNN have comparable capabil-
ity for learning the degradation process of the drawings. In addition, the kNN
method generally performs better than the CNN for color reconstruction, while
the CNN performs better than the kNN for estimating textural information.
This work presented here is based on the assumption that the degradation
between test images and target images is in the same way as in the training
images. However, in practice, different drawings and their reproductions were
made in different reproduction methods and with different degradation level.
Thus, in this work we illustrate the use of learning algorithms to learn degra-
dation process between sub-images of a drawing and its reproduction, and later
test the estimation performance via different sub-images of the drawing and its
reproduction. Although the current research work in this paper considers only
one drawing and one of its reproductions, it can also be explored to estimate
332 Y. Zeng et al.
the original drawing by using multiple reproductions and when more research
results about the instruments of drawings from other side, such as chemical, are
available. On the basis of the composition of purple inks in some of Van Goghs
drawing [8], it is possible to reproduce representative purple inks according to
contemporary 19th century recipes. In an experimental context Van Gogh like
drawings with this new ink can be made and with them the effect of exposure
to light up to 60 museum years can be simulated [9,10]. This approach can lead
to a set of reference images: original and several discolored ones (i.e. before and
after aging). This reference set can be used for the generation of a mathematical
model for fading and color changes. The inverse model can then be applied to
the discolored drawing.
The comparison experiments in this paper have shown that the estimation
performance of the simple linear regression method is very limited, since the
degradation process is not linear. Comparable estimation results were obtained
by using the kNN and the CNN. The estimation in the CNN is based on both
local and global information of the drawing, while the kNN is based on com-
paring the distance between pixel-level features and predicts the target image
pixel by pixel. Furthermore, to obtain sufficient estimation accuracy, the kNN
method should use all pixels in training images, which will lead to extremity
high computational complexity. Thus, due to limited computational time, we
use approximately one tenth training data in the kNN method in this work. As
a matter of fact the estimation performance of kNN can be improved by using
more training data. Another possible approach to improve the estimation perfor-
mance of kNN and reduce its computational complexity is to use well constructed
features, such as scale-space features [11], instead of raw pixels value. Moreover,
since the CNN is good at using local and global information for estimation, while
the advantage of the kNN is using detail information of local pixels, for future
research work it will be interesting to investigate how to combine the CNN with
the kNN to further improve color reconstruction performance.
Acknowledgments. This research is a part of the REVIGO project, supported by

the Netherlands Organisation for scientific research (NWO; grant 323.54.004) in the
context of the Science4Arts research program.
References
1. Vellekoop, M., Jansen, L., Geldof, M., Hendriks, E., de Tagle, A.: Van Goghs Studio
Practice. Van Gogh Museum, Brussels (2013)
2. Edwards, A.L.: Multiple Regression and the Analysis of Variance and Covariance.
W.H. Freeman, San Francisco (1979)
3. Devroye, L., Gyrfi, L., Lugosi, G.: A Probabilistic Theory of Pattern Recognition.
Springer, New York (1996)
4. Loog, M., van Ginneken, B., Schilham, A.M.R.: Filter learning: application to
suppression of bony structures from chest radiographs. Med. Image Anal. 10(6),
826–840 (2006)
5. Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale
image recognition. In: Proceedings of ICLR 2015 (2015)
6. Zhou, B., Lapedriza, A., Xiao, A., Torralba, J., Oliva, A.: Learning deep features
for scene recognition using places database. In: Proceedings of NIPS 2014 (2014)
7. Ren, S., He, K., Girshick, R., Sun, J.: Faster R-CNN: towards real-time object
detection with region proposal networks. In: Proceedings of NIPS 2015 (2015)
8. Neevel, H., van Bommel, M., Vellekoop, M., Lingbeek, N.: From purple to brown
analysis of Van Goghs drawing inks; non-invasive identification of inks on Van
Goghs drawings and letters. In: Proceedings of the 15th Triennial Conference, vol.
15, pp. 278–284 (2008)
9. Geerts, J.: Light damage on paper objects with coloured media; towards a differ-
ential sensitivity classification, MA thesis (2010)
10. Reissland, B., Geerts, J., Confortin, D., Neevel, J.G., van Velzen, B., Ligterink, F.:
Deep purple? Visualizing the alarming low lightfastness of crystal violet inks. In:
XIIth Congress of the International Association of Book and Paper Conservators,
p. 37, August 2011
11. Van Vliet, L.J., Young, I.T., Verbeek, P.W.: Recursive Gaussian derivative filters.
In: Proceedings of ICPR 1998 (1998)
Full Paper: Digital Cultural Heritage in
Education, Learning and Training
“Human” Technology in the Digital Era: Freehand
Images and Analysis of Cultural Heritage –
The Know-How and Its Applications
Anna Lobovikov-Katz ✉
( )
Faculty of Architecture and Town Planning, Technion - Israel Institute of Technology,

Technion City, 32000 Haifa, Israel
[email protected]
Abstract. Rapidly developing advanced methods and techniques often displace

the traditional ones. But might such “traditional” perception of the “old” as hope‐
lessly outdated cause us to overlook its intrinsic qualities? Could a relevance for
actual disadvantages be suggested, could a traditionally time-consuming tech‐
nology be transformed into an effective one, with its original values preserved?
This paper reconsiders the role of freehand sketching in modern conservation of
cultural heritage by shifting the main focus from the result to the process. It
presents a method, combined with the rapid learning methodology for achieving
this traditional artistic ability, and examines its focused application to the visual
analysis of cultural heritage by non-artist users. This paper demonstrates its wide
accessibility to the general public and conservation experts, and examines its uses
in modern multi-and interdisciplinary conservation of built heritage through
recent results of the application of the method in national and international
projects.
Keywords: Freehand sketching · Visual analysis · Conservation of cultural

heritage (CCH) · Learning methodology · Documentation · Cultural heritage (CH)
1 Introduction: Structure of the Problem
Freehand images are traditionally associated with arts and art-related areas. Sometimes,
historic images are used as an auxiliary source of information about historic monuments.
However, active use of conservation-focused modern freehand image is not part of the
methods presently applied to the contemporary conservation of cultural heritage. Exclu‐
sion of freehand images from this field might be understandable, due to several common
assumptions:
(A) The use of freehand methods cannot produce precise and detailed results, compared
to images produced by the use of high-technology techniques;
(B) Producing a freehand image is traditionally thought to be time consuming;
(C) Acquiring an acceptable skill in freehand sketching requires much time, effort and
aptitude, which reduces its application to a very limited number of experts.

DOI: 10.1007/978-3-319-48496-9_27
338 A. Lobovikov-Katz
Are these A-B-C assumptions real? Can a proper solution be found to overcome the
difficulties? Let us cast a fresh analyzing glance at the role of freehand sketching in the
modern conservation of cultural heritage (CCH), and focus on the following questions
which this paper addresses:
• What uses of freehand sketching and freehand images are relevant in the contempo‐
rary highly-technological field of conservation of CH;
• How can they be applied;
• Who can benefit from CCH sketches and sketching?
Actually, the above A-B-C assumptions focus on the outcome qualities, while they
miss its purposes. In this paper, the sketch (image) is approached as part of the sketching
process.
The paper presents the method of “Focusing the Eye” (FtE) on conservation-related
data of each specific CH object. Combined with “Rapid Learning Methodology in Free‐
hand Sketching” (RaLeMeFS), it can serve as a powerful tool for the analysis of CH
objects for conservation purposes. The approach is based on methods initially devel‐
oped, applied and introduced by the author in 2002, 2003, 2009 [1–3].
[FtE] + [RaLeMeFS ] = [ Visual Analysis Tool for CCH through Sketching]

The paper presents the results of recent development and application of the above
methods in national and international projects. It demonstrates the easy applicability of
the method, and exemplifies the results of its application by diverse types of users who
had little or no experience in arts before the application period.
2 WHAT: Freehand Image as an Engine of Conservation Analysis
In the past, freehand sketching and drawing were an integral part of architectural educa‐
tion, and they were traditionally applied in conservation of built heritage [4, 5]. Archi‐
tectural curricula produced the main heritage conservation expert force—conservation/
restoration architects, and included a significant number of courses in drawing, painting
and sculpture, with many sources available [6–8]. Since the closing decades of the 20th
century, the important development of advanced methods and techniques for collection,
processing and production of visual and multi-dimensional data has led to a wide
replacement of freehand techniques in both architectural and engineering education, and
in the field of recording and survey of historic buildings. At the same time, the under‐
standing of specific and unique values of freehand applications for architects, engineers
and scientists is gaining more attention, and sometimes originates from the ICT
community [9]. The role and uses of sketching as an instrument for creative synthesis
in various areas of design have been profoundly studied over the decades [10–13].
Moreover, recent developments in sketch-based modeling are opening new opportuni‐
ties of integration of freehand sketches as free input in computer-aided design [14–16].
However, the majority of architects and engineers seem to no longer possess the freehand
drawing ability. Some developments, aimed at providing a short path to successful
“Human” Technology in the Digital Era 339
drawing, are mostly addressed to general public [17], or non-conservation architects

[18], but they don’t have heritage conservation in mind.
In the context of CH conservation, the production and application of freehand images
are commonly seen as fossils, with unchangeable properties. However, if we approach
them as flexible and goal-adjustable, the spectrum of their uses may be extended signif‐
icantly. This also allows reconsidering the A-B-C Assumptions.
With regards to Assumption B: producing a freehand image is time consuming; -
indeed, it takes far more time to produce a free hand image than e.g. to take a photograph.
But what are the aims of each action? Freehand drawing and precise imaging, produced
with the aid of advanced technologies, differ by their driving forces. While advanced
technologies provide a wide range of methods and equipment for data collection and
multi-layer analysis, usually by means of pre-developed processes, a freehand image
has a first-hand connection with the brain of its human producer. A freehand image is
flexible and is produced through the immediate contact, reflection and reaction to the
experience, perception, knowledge and understanding of its maker in real time (Fig. 1).
Fig. 1. CCH - targeted sketching process visual analysis © A. Lobovikov-Katz
In conservation of CH, a freehand image is more useful for analysis and decision
making than for the accumulation of precise data. In many cases it is simply unnecessary
to produce a detailed image which can be obtained by the use of advanced technologies.
A sketch serves as a self-focusing instrument for its producer. Such a sketch should not
be significantly time-consuming. In this context, the terms “image” and “produce” lose
their original role, because the main focus here is shifted from the result to the process
of understanding and analyzing. This is the major contribution of freehand sketching to
CCH.
With regard to Assumption A: the use of freehand methods cannot produce accurate
and detailed results, - accuracy can be achieved when a sufficient level of proficiency
in sketching is achieved through the learning process [19]. Now, let us focus on Assump‐
tion C which claims that acquiring an acceptable skill in freehand sketching requires
much time, effort and aptitude, which reduces its application to a very limited number
of experts. The major aim of “Rapid Learning Methodology in Freehand Sketching” is
precisely to do this - to shorten the duration of the learning period needed to attain a
sufficient level in freehand sketching. In this way, it allows practically everybody to
quickly acquire a sketching ability, and open up sketch-powered visual analysis to wide
range of users within and beyond the heritage conservation community.
Different types of conservation-focused sketchers might use freehand sketches in
different ways. Conservation experts are capable of immediately understanding the
conservation implications of their discoveries made through sketching, and the major
outcome might be operational conclusions, rather than the sketch itself. However,
sketchers who are not trained in conservation, but who produce a sketch with a specific
focus on its contribution to CCH, cannot achieve full-scale conclusions resulting from
their sketching process. In this case, the process of understanding which occurs during
their sketching can help them produce the result - the sketch, - which should contain
enough information to make it useful for its subsequent analysis by a conservation
expert. In this case also, the sketch presents a visual result of a preliminary analysis
purposefully focused on conservation by a sketcher who is not a conservation expert
[20] (Fig. 2).
Fig. 2. Duality of sketching result: the image and the process. Direct and indirect use of sketching
in CCH by conservation experts © A. Lobovikov-Katz
As demonstrated in the course of different educational heritage studying frameworks,

sketching can be used to a varying extent and purpose at different levels of CH expertise,
but they have a common basis - visual analysis, powered by the sketching process,
activating deeper levels of understanding than by merely looking at a monument.
3 HOW: Focusing the Eye - CCH-Focused Visual Analysis

in Sketching
Two skills are required from a sketcher for a successful representation for CCH:
• to perform an educated visual analysis of CCH objects:
• to produce an image of a CCH object.
This section focuses on the first skill as part of the sketching process - the ability to
understand the visual data of a heritage object perceived by the sketcher. As shown in
the European project ELAICH (Educational Linkage Approach In Cultural Heritage)
that was carried out in the framework of the Euromed Heritage Programme 4 (2009–
2012) - ELAICH ENPI-2008/150-583, it is possible to educate in a relatively short time
an audience, non-trained in CCH, - to achieve a good level of understanding of CCH
basic problems, by the use of “ELAICH Methodology” [21]. However, to enable a
sketcher to produce a sketch of conservational analytical value, it is necessary to train
his eye and mind to capture the multilayered data on site. To become effective for actual
sketching, CCH focus should be formulated as simply as possible. At the same time,
focus definition should not overlook the essence of CCH. This includes the efforts to
understand and preserve a unique complex of values of each heritage object; the efforts
on its recovering (to possible limits) from deterioration and preventing further decay,
through respect to authenticity, taking into consideration methodologies, principles and
the development of criteria. In this context, CCH is formulated as reaching a balance
between the values (positive) and the deterioration (negative) of CH objects (Fig. 3).
This simplified formula is the basis for focusing the Eye (FtE) of non-trained sketchers
and it can be of use to both CCH experts and the general public in the real time of in-
situ sketching.
Fig. 3. Focusing the eye: capturing the characteristics of both the values and the deterioration of
CH © A. Lobovikov-Katz
While conservation experts are highly knowledgeable in their CCH area, they still
need to sharpen their skills in the unfamiliar area of CCH sketching. It is not easy for
an untrained-in-the-art sketcher to capture and transmit to/by sketch many of the visual
characteristics of a sketched object. This might be caused not merely by the complexity
of the sketching process, but also in the difficulty to combine analytical thinking with
creative processes. Furthermore, experts from many areas of science and technology are
involved in the contemporary conservation of cultural heritage, and some of them do
not deal with the analysis of historic buildings or sites on a daily basis, e.g. experts in
analytical chemistry need to develop different skills for sketching than those required
in their laboratory.
For the general public, who are not conservation experts, such guidance is indis‐
pensable. For them, it is necessary to go further, and get more into details, such as
learning to perceive an entire edifice as an entity combined of its integral parts: geometry
(shape, space) and materials [22]. Both geometry and materials play an active role in
solving contradictions frequently encountered on historic monuments, such as between
the urge to preserve an architectural value of a building or a detail, versus its poor
physical state. The difficulty of conservation decisions and interventions may be rooted
in the fine geometry of an authentic detail and the inability to preserve it due to lost
geometry of its present state, while geometry is lost through the loss of material, and
possibly - the loss of structural stability. Developing awareness and understanding of
such problems sharpens and educates the visual focus of general public sketchers. Their
efforts in actual sketching are based on this solid intellectual basis - not just looking at
a building but to know where to look and what to look for.
In Fig. 4, macro geometry relates to a large-scale built heritage object, - the entire
architectural ensemble, structure, building or its large parts. Micro geometry relates to
elementary single units, e.g. a single stone, column capital, etc., with attention to both
Fig. 4. Focusing the eye: capturing the characteristics of both the values and the deterioration of
CH - between geometry, material and environment © A. Lobovikov-Katz
material properties and material deterioration. This simple scheme demonstrates the
interconnection between environment, geometry and material, with regard to both the
values and the deterioration, e.g. the original architectural choice of building material
depends on its availability, and at the same time its original material properties prede‐
termine its durability under specific environmental conditions.
“Focusing the eyes” of CCH sketchers should be undertaken before the start of their
actual on-site sketching of historic buildings. This should be combined with introducing
the actual methodology for freehand sketching based on the author’s Rapid Learning
Methodology in Freehand Sketching.
4 WHO: Accessibility of the Method, Its Uses and Application

Examples
In order to sketch, a basic proficiency in freehand sketching is needed. The method is

accessible to practically everybody, through the Rapid Learning Methodology in Free‐
hand Sketching. This section of the paper presents several examples of its application
combined with FtE - “focusing the eye” on CCH. Different types of participants, from
high school students to conservation experts took part in the sessions given by author,
in the framework of several projects. The majority of all participants had no experience
in freehand sketching before the sessions. Duration of learning-sketching sessions was
different for each group, as was the relative duration of introducing the sketching meth‐
odology (RaLeMeFS) and the “focusing the eye” on CCH.
4.1 Case Study 1: The Use of CCH Freehand Sketching by CCH Experts
This example demonstrates the use of CCH freehand sketching in the course of a pilot
learning experience undertaken in sessions on “Understanding cultural heritage through
sketching – rapid learning methodology” as part of the Training School, organized by
COST i2MHB and the Fundación Santa María la Real del Patrimonio Histórico on its
premises in Aguilar de Campoo in Spain, January 27–29, 2016.
As part of the Training School, sketching sessions (2 h lecture and demonstration;
home work; 2.5 h practical application on a historic site; 1 h conclusion and analysis of
results) were given by the author to a group of experts, MSc and PhD students from the
following areas of CH conservation: history, archaeology, architecture, art history, soci‐
ology, cultural heritage, conservation of cultural heritage, civil eng., building pathology,
chemistry, electrical eng., economics, management. Some participants came from a
combined educational background in several areas, i.e. BSc., MSc., and PhD in different
fields. Only a small number of participants were architects.
At the conclusion of the sessions, participants filled in questionnaires. Of 22 partic‐
ipants, 8 had no experience in freehand drawing, sketching or painting before the
sessions, 2 had “almost none” or “very little” experience (45 % altogether), while 12
participants had experience in this area.
There was very little difference between the two types of participants with regard to
the main difficulties during the sessions. As reported by the participants, many
difficulties were related to capturing “perspective” and “proportions”. 50 % of the

participants with no or very little experience in sketching wrote that “everybody can
learn to draw”, as an answer to their summing up of the session.
72 % of all participants answered that sketching is useful for every expert in this
field. Majority of participants were interested in further learning of sketching. 86 % of
all participants were planning to use sketching in their main field of work. According to
participants’ answers, a wide range of possible general uses of sketching in conservation
of cultural heritage could be possible. The participants provided interesting and often
unusual answers. One answer expresses the essence and relevance of sketching for
understanding and preserving of cultural heritage. This was provided as an answer to
the question: Q: “What can be the use of sketching in conservation of cultural heritage
in general (if any?)”. A: “To sketch what you do not see”. This quotation puts in a nutshell
the great value of sketching as a tool for deepening the understanding and discovering
the data which is often “invisible” to a non-sketching observer.
Fig. 5. Sketches, produced by two conservation experts - participants of practical session on

“Understanding cultural heritage through sketching”, as part of the COST i2MHB Training
School, Fundación Santa María la Real del Patrimonio Histórico, Spain, 2016: FtE and RaLeMeFS
were applied. Both participants (D.Sc., construction economist (left) and MSc, building engineer;
restorer (right)) had no previous experience in sketching, painting or freehand drawing before the
session. Nevertheless, the sketches demonstrate accurate graphic analysis of geometry and
material, including visible diverse degrees of deterioration.
After the lecture and demonstration, during the practical part of the session at the
historic site, the participants were asked to pick a detail or a small portion of a wall or
structure on which to focus their sketching. In spite of the short duration of the practical
session, some participants without previous experience in sketching showed very good
results. Some sketches of “newcomers to the art” demonstrated attentive visual analysis
of conservation problems of a historic monument. At first, many participants experi‐
enced difficulty with proportions and perspective. However, many quickly grasped the
rules and discarded their mistakes already by the end of the 2-hour sketching session.
At the same time, they apparently enjoyed “understanding heritage through sketching”,
which possibly helped them to persist until completing their task in the severe winter
cold of Northern Spain. To conclude this specific experience, the main value of CCH
sketching by conservation experts is “fueling” the process of their analytical under‐
standing rather than the production of a detailed sketch. At the same time, FtE-aided
sketcher-sketch analytical dialogue should be based on an acceptable level of sketching

which can be achieved by RaLeMeFS (Fig. 5).
4.2 Case Study 2: The Use of CCH Freehand Sketching by General

Public - The Case of High School Students
This example presents a very short session given by author to a group of some 25 students
of the Ben Zvi high school at Kiriat Ono, with the support of and at the premises of the
Society for Preservation of Israel Heritage Sites (SPIHS) at the Mikveh Israel School
Visitor Centre in Tel Aviv, Israel, in May 2016.
The students had several introductory meetings to the conservation of cultural
heritage provided by the SPIHS experts before this session. They had no previous expe‐
rience in sketching. The session duration was about 1.5 h. It included the “Focusing the
Eye” part and a short review of different types or projections; a very brief explanation
and demonstration on sketching methodology (RaLeMeFS); and about 20 min of prac‐
tical sketching of historic buildings or their elements. Introduction to and practicing
RaLeMeFS were too short (about 40 min altogether) to prevent mistakes in proportions
and perspective, which affected negatively the reliability of many sketches.
However, students’ choice of their sketching subjects demonstrated their ease and
success with using FtE. Right from the start of the practical part of the session, as they
stepped from the classroom into the historic compound, they “focused their eyes“ on
conservation problems and values, e.g.: artistic values and material deterioration of
specific architectural elements, material deterioration problems of plaster; cracks in
buildings and supporting walls; etc (Fig. 6).
Fig. 6. Fast on-site sketch by a high-school student at a practical part of a lesson on understanding
historic sites through sketching: material deterioration of a ground storey corner of a historic
building, and pavement.
4.3 Case Study 3: The Use of CCH Freehand Sketching by Non-conservation

Architecture Students
These architecture students had no previous course in conservation. Some of them took
a course in architectural drawing assisted by RaLeMeFS. Freehand sketch demonstrated
here was produced in the framework of the course in documentation and survey of
historic buildings and sites. The course was given by the author at the Faculty of Archi‐
tecture and Town Planning, Technion – Israel Institute of Technology in Haifa. It
consisted of lessons in a university classroom, exercise in Haifa, and three to four days
of precise measurements on a World Heritage site in Acre (Akko) with the help of simple
equipment supplied by the Transportation and Geo-Information Engineering section of
the Faculty of Civil Engineering. This course had a strictly precise orientation. At the
end of the course, students had to produce several layers of mapping according to guide‐
lines. Freehand sketching was not required. Furthermore, CAD disciplines dominate the
faculty curriculum, with a very limited number of freehand-related, not obligatory
courses. Nevertheless, the few participants who succeeded to acquire basic sketching
skills before this course used sketching to the benefit of their documentation teams.
Furthermore, the sketch of a heavily deteriorated column capital, presented here, helped
Fig. 7. Sketch from a field book, part of the teamwork in students’ documentation and mapping
project, in the course “Documentation of historic building and sites” Faculty of Architecture and
Town Planning, Technion, (supervised by the author) 2011. Sketch by Ori Roll. This freehand
sketch focuses on severe deterioration of the column’s capital. It also points out the initial visible
stage of detachment of a stabilizing metal brace beneath it; and hints at material deterioration of
an arch above the capital, on the left. Basing his opinion on this sketch, a conservation expert
might recommend urgent engineering investigation of such details of the courtyard in order to
examine their material deterioration and structural stability altogether. In the specific context of
this students’ project, the sketch helped them to discover and analyze the main natural deterioration
cause on this part of the monument.
them to formulate the right questions and reach correct conclusions about the main
deterioration causes, their geometrical distribution on the monument, and relative dating
of the capitals of the courtyard (Fig. 7).
As demonstrated by these examples, on-site sketching brings the sketching person to
deeper levels of understanding than by merely looking at the monument. This is true for
both high school students and conservation experts, though their use of sketching differs.
5 Conclusions
Human visual system is able to process complex data, including information on both
geometry and materials of complex objects. Combining this natural ability with a focus
on conservation data and problems of an observed cultural heritage object with the help
of FtE (Focusing the Eye), and at the same time activating the process of visual analysis
through sketching, can be an invaluable aid to the conservation of cultural heritage. FtE,
supported by RaLeMeFS, allows for the inclusion of traditional freehand sketching into
modern interdisciplinary and multidisciplinary conservation research and practice. Due
to its simplicity, the method can be used by the general public and therefore enable large
numbers of historic monuments to be studied and monitored regularly on a basic level.
The results of such wide active visual observation can be used by conservation experts.
Furthermore, conservation experts can use this method directly, to effectively enhance
their process of visual analysis of historic buildings and sites under study or conserva‐
tion. This method can be also applied by other heritage-related communities, who have
become more involved in conservation in recent decades, e.g. archaeologists.
Activating visual analysis by FtE and RaLeMeSF, successful results of involving
e-learning in the education of the general public for CCH [23–25], and the rise of interest
and achievements of the digital community in the development of sketch-based
modeling in the recent decade [26] add new dimensions to an old traditional, two-
dimensional technique of freehand sketching.
Acknowledgements. The author would like to express her appreciation to all organizers and
participants of training and learning activities, for their support and enthusiasm, especially COST
TD1406 (I2MHB) - Innovation in the intelligent management of heritage buildings; Fundación
Santa María la Real del Patrimonio Histórico, Aguilar de Campoo, Spain; Society for Preservation
of Israel Heritage Sites (SPIHS); Technion-Israel Institute of Technology.
References
1. Lobovikov-Katz, A.: A new teaching methodology in freehand drawing and painting and its
application to the field of conservation of monuments. In: Galan, E.M., Zezza, F. (eds.)
Protection and Conservation of the Cultural Heritage of the Mediterranean Cities, Proceedings
of the 5th International Symposium on the Conservation of Monuments in the Mediterranean
Basin, Sevilla, Spain, 5–8 April 2000, pp. 593–597. Swets & Zeitlinger, Lisse (2002)
2. Lobovikov-Katz, A.: Heritage education for heritage conservation (contribution of
educational codes to study of deterioration of natural building stone in historic monuments).
Strain Int. J. Exp. Mech. 45(5), 480–484 (2009)
3. Lobovikov-Katz, A.: Renovation approach to the role and methodology of teaching graphic
disciplines as an integral part of education of architects for conservation (comprehensive
teaching program approach). In: Proceedings of the Scientific Congress: 7th International
Symposium of the Organization of World Heritage Cities, (Keeping Heritage Alive –
Education and Training for the Preservation of Cultural Heritage), Rhodes, 23–26 September
2003 (2003)
4. Feilden, B.M.: Conservation of Historic Buildings. Butterworth Scientific, London (1982)
5. Burns, J.A.: Recording Historic Structures: Historic American Buildings Survey, Historic
American Engineering Record. American Institute of Architects Press, Washington, D.C.
(1989)
6. Alberti, L.B.: Ten books on architecture. The Academy of Architecture of the USSR, Moscow
MCMXXXV (Лeoн Бaттиcтa Aльбepти: Дecять Книг o Зoдчecтвe, Издaтeльcтвo
Bcecoюзнoй Aкaдeмии Apxитeктypы, Mocквa) (1935)
7. Vasari, G.: Vasari On Technique: Being the Introduction to the Three Arts of Design,
Architecture, Sculpture and Painting, Prefixed to the Lives of the Most Excellent Painters,
Sculptors and Architects. Dover, New York (1960)
8. Tichonov, S., Demijanov, V., Podrezkov, V.: Drawing. Moscow, Stroiizdat (Tиxoнoв, C.B.,
Дeмьянoв, B.Г., Пoдpeзкoв, B.Б.: Pиcyнoк, Mocквa Cтpoйиздaт) (1983)
9. Duhovnik, J., Demsar, I., Drešar, P.: Technical freehand sketching, Chap. 2: In: Duhovnik,
J., Demsar, I., Drešar, P. (eds.) Space Modeling with SolidWorks and NX, vol. XIV. Springer,
Cham (2015)
10. Goldschmidt, G.: The dialectics of sketching. Creativity Res. J. 4, 123–143 (1991)
11. Lane, D., Seery, N., Gordon, S.: A paradigm for promoting visual synthesis through freehand
sketching. Des. Technol. Educ. 15(3), 2010 (2010)
12. Verstijnen, I.M., van Leeuwen, C., Goldschmidt, G., Hamel, R., Hennessey, J.M.: Sketching
and creative discovery. Des. Stud. 19(4), 519–546 (1998)
13. Goldschmidt, G.: On visual design thinking: the vis kids of architecture. Des. Stud. 15(2),
158–174 (1994)
14. Olsen, L., Samavati, F., Sousa, M.C., Jorge, J.: A taxonomy of modeling techniques using
sketch-based interfaces. Eurograph. State Art Rep. 1(1–4), 1 (2008)
15. Kondo, K.: Interactive geometric modeling using freehand sketches. J. Geom. Graph. 13(2),
195–207 (2009)
16. Olsen, L., Samavati, F., Sousa, M.C., Jorge, J.: Sketch-based modeling: a survey. Comput.
Graph. 33, 85–103 (2009)
17. Edwards, B.: The New Drawing on the Right Side of The Brain. Tarcher/Putnam, New York
(1999)
18. Yee, R.: Architectural Drawing: A Visual Compendium of Types and Methods. Wiley, New
York (1997)
19. Lobovikov-Katz, A.: Freehand sketching in the contemporary conservation of monuments,
water and cultural heritage. In: Zezza, F., Perthuisot, V., Plancon, A. (eds.) Proceedings of
the 7th International Symposium on the Conservation of Monuments in the Mediterranean
Basin, Orleans, France, June 2007 (2008)
20. Lobovikov-Katz, A., Moropoulou, A., Konstanti, A., Calderón, P.O., Grieken, R., Worth, S.,
Cassar, J., Angelis, R., Biscontin, G., Izzo, F.C.: Tangible versus intangible in e-Learning on
cultural heritage: from online learning to on-site study of historic sites. In: Ioannides, M.,
Magnenat-Thalmann, N., Fink, E., Žarnić, R., Yen, A.-Y., Quak, E. (eds.) EuroMed 2014.
LNCS, vol. 8740, pp. 819–828. Springer, Heidelberg (2014). doi:10.1007/978-
3-319-13695-0_84
21. Lobovikov-Katz, A., Konstanti, A., Labropoulos, K., Moropoulou, A., Cassar, J., Angelis, R.:
The EUROMED 4 project “ELAICH”: e-Tools for a teaching environment on eu
mediterranean cultural heritage. In: Ioannides, M., Fritsch, D., Leissner, J., Davies, R.,
Remondino, F., Caffo, R. (eds.) EuroMed 2012. LNCS, vol. 7616, pp. 710–719. Springer,
Heidelberg (2012). doi:10.1007/978-3-642-34234-9_75
22. Lobovikov-Katz, A.: The correlation between the technological and conservation aspects and
those of urban appearance in a stone-built environment: an evaluation approach, Chap. 4.2.
In: Kourkoulis, S. (ed.) Fracture and Failure of Natural Building Stones – Applications in the
Restoration of the Ancient Monuments, pp. 201–215. Springer, Dordrecht (2006)
23. Lobovikov-Katz, A.: The virtual and the real: e-learning in interdisciplinary education – the
case of cultural heritage. In: The 13th Annual MEITAL National Conference “New
Technologies and Their Evaluation in Online Teaching and Learning”, Technion - Israel
Institute of Technology, Haifa, June 2015. http://meital.iucc.ac.il/conf2015/papers15/
A3_3.pdf
24. Lobovikov-Katz, A., et al. (eds.): ELAICH (Educational Linkage Approach in Cultural
Heritage) Manual. Technion (2012). http://www.elaich.technion.ac.il/e-learning
25. ELAICH e-learning platform. http://www.elaich.technion.ac.il/e-learning
26. Kazmi, IK., You, L., Zhang, J.J.: A survey of sketch based modeling systems. In: 11th
International Conference on Computer Graphics, Imaging and Visualization (CGIV) (2014)
Adult and Children User Experience with Leap Motion
in Digital Heritage: The Cycladic Sculpture Application
Panayiotis Koutsabasis ✉ and Spyros Vosinakis

( )
Interactive Systems Design Lab, Department of Product and Systems Design Engineering,
University of the Aegean, Syros, Greece
{kgp,spyrosv}@aegean.gr
Abstract. Recent advances in low-cost sensor technologies, such as Microsoft

Kinect and Leap Motion allow kinaesthetic interactions with interactive 3D
applications. Museums and heritage institutions can significantly benefit from
kinaesthetic applications that provide a more experiential approach for learning
about cultural heritage; however detailed evaluations of the user experience are
still scarce. This paper presents the development and user-centred evaluation of
a cultural heritage application about sculpturing Cycladic figurines, which places
users in the role of an ancient craftsman or sculptor who progressively creates a
statue by selecting and applying the appropriate tools with bare-hand interactions
tracked by the Leap Motion sensor. The evaluation of the user experience of ten
adults and ten children in two subsequent studies reveals that users find the expe‐
rience very positive and engaging but usability and tracking issues remain. We
identify these issues and propose design guidelines to address them.
Keywords: Cultural heritage · Kinaesthetic interaction · User experience · Adults ·

Children · Concurrent think-aloud · Usability testing · Leap Motion
1 Introduction
The dissemination of cultural heritage has gained significant benefits from digital tech‐
nology. Various digital applications related to both tangible and intangible cultural
heritage have been made available to the public, which rest on various established trends
in entertainment software [1], such as Virtual Worlds [16], Serious Games, Interactive
Storytelling, Location-based Games, Augmented Reality, etc.
Recent advances in sensor technology have led to novel low-cost devices that can
track human body motion which allow the user to make use of their body movements
and gestures to interact with digital content. Notable examples are the Microsoft Kinect,
which can track the position and movement of the whole body and recognize user
gestures and the Leap Motion which tracks the placement and motion of the hands only.
The possibility of kinaesthetic interactions may expand the potential of digital
heritage applications into new paradigms. The kinaesthetic approach is considered an
important pedagogy for museum education [5], especially for highlighting intangible
aspects of cultural heritage, e.g. habits, rituals, everyday activities of the past, etc. Addi‐
tionally, gesture-based interactions are generally more suitable input methods for 3D

DOI: 10.1007/978-3-319-48496-9_28
Adult and Children User Experience with Leap Motion 351
digital heritage environments because they involve natural motion in three dimensions.
Last but not least natural and intuitive interactions may make the user experience more
fun and, as such, retain user interest and engagement for a longer period of time [6].
Up to now, there are limited digital heritage kinaesthetic applications, but there
seems to be a growing interest in this direction of research and development. In addition,
the impact of kinaesthetic heritage applications on the visitor user experience is largely
unknown. There are too few reasonably detailed studies of their usability and impact on
the user experience, which poses the question of whether kinaesthetic digital heritage
applications provide valuable and usable experiences to users.
In this paper we present the design, development and user-centred evaluation of a
kinaesthetic digital heritage application that provides users with simplified virtual
sculpting of Cycladic figurines which is exercised by applying respective gestures with
their bare hands. We also present the results of a detailed evaluation of the user expe‐
rience with the participation of ten adults and ten children in two subsequent studies that
included both formative and summative assessments based on the concurrent think-
aloud protocol, performance measurements and a short questionnaire. The evaluation
results indicate that users gain an engaging and educational user experience through
kinaesthetic interactions, however various usability and tracking issues remain.
2 Background and Related Work
Kinaesthetic or embodied interaction [2] is about exploiting any type of user movements
to interact with a computer, a distant display, projected content or a technologically
enhanced room or environment. This is a distinguishable style of natural human-
computer interaction which has materialized in further areas including: whole-body
interaction [3], which refers to the use of body movements and postures; mid-air inter‐
action [7], which refers to the use of (whole) hand gestures in the mid-air in order to
interact with distant displays; and bare-hands interaction [14], which refers to the use
of the hand palms and fingers mainly for desktop or kiosk-oriented interactions.
Kinaesthetic interaction has not yet substantially developed into a considerable
corpus of applications in cultural heritage. In [11], a number of large-scale virtual reality
installations of whole-body interactions with projected 3D content in museums are
presented, supported by the Kinect sensor. A user can made simple moves into the
physical space to control her avatar inside an Etruscan grave. A user could also apply
mid-air gestures to interact with digital objects, but would need to learn a gesture
vocabulary.
The installation of [4] comprises of a Kinect-based installation with which a user
may navigate in restored sites of ancient Rome, and a kiosk-based application in which
another user may inspect digital artefacts with bare-hand interactions supported by the
Leap Motion. The evaluation results are encouraging, but it is reported that users
required to guess for the appropriate manipulations regarding the kiosk-based applica‐
tion.
In the Leap Motion app store [8], the applications related to cultural heritage are
underdeveloped. From the total of 230 apps, 99 are free of charge and only one is about
352 P. Koutsabasis and S. Vosinakis
cultural heritage, ‘TomBraining, the Gallery’, which offers a virtual tour to a collection
of 250 paintings. Users use the Leap motion controller to travel in the virtual world using
respective gestures.
3 The Cycladic Sculpture Application
3.1 Main Sources of Information
The process of Cycladic figurine creation has been documented on an illustrative book
regarding the Cycladic sculpture process officially distributed by the museum of
Cycladic art [10]. We have made interviews with the author who is the curator of the
respective exhibition at the museum of Cycladic art as well as field visits in the museum.
Respective information is also provided in the museum’s Web site [9].
3.2 User Tasks and Challenges
The Cycladic sculpting application places the user in the role of the ancient craftsman
or sculptor. When the program starts, the user finds himself immersed in the sculptor’s
workspace and can see his ‘virtual hands’ i.e. the 3D representation of their hands in the
scene. Then, the user has to go through three main stages of constructing the figurine,
by selecting the appropriate tool and imitating the task of using it (Fig. 1). Throughout
the process the user sees the sculpture in a 3D form starting from a piece of marble and
slowly taking the form of the figurine.
(a) (b) (c)
Fig. 1. Screenshots of Cycladic sculpture application. A girl user performs (a) carving, (b)
smoothing, (c) engraving.
Each of the identified stages has two phases: the selection of the tool and its appli‐
cation with corresponding gestures. In the selection phase, the user has to identify the
correct tool that matches the requested task. By lowering his hand, the camera is rotated
down, and he can see the three tools arranged in front of the statue (Fig. 1a). When he
brings the hand above a tool, it is being highlighted, and its name is displayed. To select
a tool, the user has to perform a grasping gesture with his hand. If a grasping gesture is
detected and the user’s virtual hand collides with the tool, the tool is appropriately
attached to his hand, as if he is actually holding it. If the user selects a wrong tool, the
application provides feedback explaining why the tool is not appropriate for this task.
In that case the tool automatically leaves his hand and returns to its starting position.
In the task phase, the user has to use the tool appropriately on the statue mimicking
sculpting movements with his hand in order to complete the stage. Initially, the appli‐
cation presents a message and an animated clip to explain how the user should perform
the task in that stage. Then, the user is applying the tool on the sculpture and sees the
impact of his actions on the form of the model. While performing, the user receives
feedback about his progress through a progress bar and about his errors through warning
messages and designated sounds. Finally, when the task is completed, the user is notified
with a message and a music clip that the stage is over, and the application progresses to
the next one.
The main stages of interaction are carving, smoothing and engraving.
During carving, the user has to remove the exterior pieces of marble using the emery
tool (Fig. 1a). The outline of the target shape is visualized so that the user knows which
are the correct places to hit. Using the information provided from leap motion sensor,
the system attempts to detect fast hand motion towards the statue. If the edge of the tool
collides with the marble with a relatively fast speed, the marble shows a crack, a sound
of carving stone is played and a particle system generates the effect of small pieces of
marble falling on the ground. Then, a second hit on the same piece results in the piece
detaching from the statue and falling down. If the user hits three times inside the target
shape, it is marked as an error and appropriate feedback is produced.
During smoothing, the user has to rub the exterior part of the statue with the pumice
stone, in order to transform the crude edges into more detailed surfaces (Fig. 1b). The
outline of the target shape is visualized, so the user has to rub all the pieces outside this.
To detect the rubbing process, the system checks whether the tool collides with the
exterior parts, and, if so, it has to detect quite fast hand motion. A respective sound is
played, and a particle system visually indicates the process. When a piece of marble is
being rubbed, it initially becomes semi-transparent, so that the user is informed that he
is making progress, and then it fully disappears. An error is marked if rubbing motion
is detected in parts of the statue that must not be removed for one second.
During engraving, the user has to engrave the marked areas of the statue using the
pointy obsidian tool to add finer detail in the nose, the hands and feet (Fig. 1c). The
process of engraving requires slower and more focused hand movements. Therefore, the
user perspective changes, and the program zooms in the statue, so that the hand motions
can be mapped to smaller movements. The process is split in three parts: the face, the
body and the legs. When the user completes a part, the game progresses to the next one,
and the camera zooms to the respective area of the statue. Each part that needs to be
engraved is a thick line segment, in some cases slightly curved. The user has to bring
the pointy edge of the obsidian tool on the segment and move the tool slowly and
repeatedly across it in order to be engraved. The segments to be engraved are designated
with blue colour to be easily recognized on the statue. Technically, each segment that
has to be removed is split into multiple (usually two or three) objects. When each of
these objects detects engraving motion, it is removed, and the engraved part of the statue
is revealed. When the engraving task is detected, a respective sound and particle system
is being visualized to give feedback that the tool is actually touching the statue. If an
engraving process outside the designated segments is detected continuously for more
than one second, the system considers it an error.
After all three stages are completed, the statue is replaced by a detailed 3D model
of an actual figurine with a similar shape and size, which makes the user understand how
the final statue would look like. The user can freely rotate the statue using hand gestures
in order to be able to carefully observe it from various angles.
3.3 3D Modelling
The application development involved a series of 3D modelling stages starting from the
final statue and resulting to the unprocessed piece of marble. Initially, the detailed 3D
model of the figurine was created by 3D scanning a replica of a known Cycladic figurine.
Then, following the generic shape features of the figurine, a second, abstract version of
the model was created by flattening the surfaces in the z-dimension (depth) and simpli‐
fying the geometry. The latter would be the final model after the three stages had been
completed. Then, we added extra geometry to re-generate its form prior to the execution
of each of the three stages. First we filled the engraved parts with pieces of marble to
generate the form of the statue before engraving. Then, we added extra pieces of marble
to the exterior surface of the statue (only in the XY-plane) to generate the rough shape
of the statue before smoothing. And finally we added a number of larger pieces of marble
to generate the original piece that surrounds the statue. All these additions have been
unified under a single transformation tree.
This 3D model serves a basis to produce the response of the marble to user actions
in each of the three stages. In each stage there are a number of objects that need to be
removed by applying each technique. Collision detection tests combined with the iden‐
tification of the tool movement allows the system to determine if the technique is
correctly applied. If successfully applied, respective feedback is triggered, and the object
is removed. The number of remaining objects divided by their total number results to
the completion rate of that stage, which is visualized in the progress bar.
The Cycladic sculpture application has been implemented in Unity Game engine
using Leap Motion Orion SDK. In its current version it supports only interactions with
the right hand and has been tested in a desktop setup. The prototype implementation is
part of a larger project intended to be installed as a public installation for children and
adults in exhibition halls related to Cycladic art.
4 Evaluation Method
4.1 Procedure
We followed a mixed evaluation approach, in two subsequent studies with adults and
children, that enabled us to identify UX issues as well as to measure user performance.
We asked users to: (a) think-aloud during familiarizing with the use of the app,
(b) perform the tasks without being interrupted, and (c) provide final remarks and fill-
in a short questionnaire. The results of the first study concerned adult users and were
documented in detail in [15]. In this paper we present the results of the second study
with children and we provide a comparative account.
4.2 Participants and Apparatus

The first study involved ten (10) users (nine men, one woman, av. age: 27); all right-
handed; seven were students of design engineering and three were faculty. Seven partic‐
ipants had limited prior experience with the Leap Motion. The second study involved
ten (10) children (seven girls, three boys, av. age: 10.4) all right handed. All were of the
fourth grade of elementary school and they were acquainted with Cycladic figurines
from their history class. One had limited previous experience with the Leap Motion.
The application was installed on a computer with i7 2.50 GHz CPU, 16 GB RAM
and nVidia GTX 850 M graphics card. The Leap motion controller was positioned in
front of the user at the centre of the desk, at a distance of approximately 10 cm from the
screen and another 10 cm from the edge of the desk. Users were seated during the test.
4.3 Measures
The measures collected in these studies were:

• UX issues: they were mainly identified during the think-aloud session, but also during
play (from observation) and from post-test remarks (self-reported). After the end of
all tests, UX issues were processed and coded: this included rewriting (so that a single
statement resulted to the same issue), identifying common issues among users and
matching issues into wider UX and usability principles.
• Spontaneous verbal comments; they were recorded during the think-aloud phase.
• Performance metrics: task time and errors, measured by the application during the
play phase. Errors were of the following types: to select the wrong tool, to apply the
tool on an incorrect area, to perform an inappropriate gesture repeatedly.
• Questionnaire responses. Both adults and children filled in the fun toolkit [12], which
is a simple and validated instrument to assess the user experience in child-computer
interaction. An additional question was added that asked users to match tools (emery,
pumice rock, obsidian rock) for operations.
5 Results
5.1 Overall Impressions

To obtain a “bird’s eye view” of the overall impression of participants for the Cycladic
sculpture app, we asked them what they thought about this application (Table 1). All
participants provided positive responses with children being more satisfied than adults.
In particular six children found the app ‘brilliant’ and the rest of them found it ‘really
good’, while there was one adult who found it ‘brilliant’, six of them found it ‘really
good’ and three ‘good’. There were no negative responses.
Table 1. Participant responses about their overall impression of the application.

What do you think about this app? Adults Children
Brilliant 1 6
Really good 6 4
Good 3 0
Not very good 0 0
Awful 0 0
Furthermore, after the test there was unanimous remembrance of all ancient tools
operations. Although this may be considered simple to learn and remember, an inter‐
esting question (that can be investigated in future work) would be if they would have
recalled these after a traditional visit to the museum and without the use of the app.
5.2 User Performance
The average time to task (Fig. 2, left) was faster for adults for carving (adults: 48.8 s,
children: 65.2 s) and significantly faster for engraving (adults: 193.9 s, children: 284 s).
However, children were faster on smoothening (children: 67.1 s, adults: 77.4 s), which
is very interesting since that most adults had some prior experience with the Leap Motion
device, while for all but one children this was a totally new experience. This can be
attributed to some extent to that adult users had some occlusion issues during smooth‐
ening caused by their hand size and placement (see later on Sect. 5.3) that children did
not face since that their hands are smaller.
Fig. 2. Left: Average time to task. Right: Average errors per task.
Another measure of performance is average number of errors (Fig. 2, right). Adults

performed with fewer errors in carving (adults: 3.6 errors; children 4.5), while children
performed with fewer errors in both smoothening (adults: 2.9, children 1.6) and
engraving (adults 3.1, children 2.7). This is also an interesting result: children may
perform with fewer errors with this technology even when they do not have prior expe‐
rience! This result may seem inconsistent provided that children performed considerably
slowly in engraving, but it is not so: we observed that children were highly focused
during engraving which eventually resulted in fewer errors.
5.3 Particular User Experience Issues

A detailed view on particular UX issues was obtained during the think-aloud session
and the post-test interview. A total of 197 verbal comments were obtained (104 from
adults and 93 from children). The analysis of the verbal comments was made as follows:
• Firstly, the verbal comments were categorized into UX issues (75 for adults, 77 for
children) and other comments (29 for adults, 16 for children).
• Then, we examined each UX issue to identify uniqueness (Table 2). Some issues
were re-written to be better reflected. A total of 36 unique UX issues were identified
(26 were unique for adults and another 26 issues were unique for children).
• Finally, we mapped UX issues into design and usability principles for 3D interactions
(Fig. 3). We identified eight (8) principles in which the identified issues fall into:
Look & Feel, Feedback, Tracking, Gesturing, User Positioning, Gamification,
Control/Reach and Visibility. These are ‘working’ categorizations to enable sense-
making, based on our experience as UX designers and researchers and not any kind
of a formal classification.
Physical setup issues severely affected the UX of the app. These include tracking
(14.7 % for adults; 20.3 % for children) and user positioning (13.3 % for adults, 6.3 %
for children). More specifically, these refer to the quality of hand tracking by the Leap
in relation to the user positioning, the placement of the controller and the screen. Most
users experienced some kind of a tracking issue, which was either that the Leap lost
hand tracking (6/10 adults; 2/10 children), or that the hand colouring changed unex‐
pectedly (3/10 adults; 5/10 children), or that the tracked virtual hand moved or took
stances different from user moves (2/10 adults; 4/10 children). All these effects occurred
for a few seconds but interrupted the user flow resulting to an unpleasant UX.
Fig. 3. Types of UX issues identified from think-aloud.
Furthermore, most users – especially adults - felt uneasy with the positioning of their
body or hands during interaction. This happened often when they needed to perform
operations at the limits of the tracking area (which generally corresponded to the edges
of the scene). Some users felt they wanted to stand up (3/10 adults, 3/10 children) in
Table 2. User experience issues identified by participants (adults and children) with occurrences
and frequency, classified into more general issues.
# U X Issue A C Freq.
1 Not good perception of depth 4 9 65.0%
2 Fatigue during engraving 5 7 60.0%
Look & Feel
3 Engraving takes too long 3 6 45.0%

4 Hand passes through the statue 2 5 35.0%
5 Not sure what the problem is (during unsuccessful operations) 2 1 15.0%
6 User requests to add music (overall) 1 2 15.0%
7 Vertical engraving seems to work better 1 0 5.0%
8 Lost hand tracking (hand disappears from scene for a while) 6 2 40.0%
9 Leap switches between right/left hand unexpectedly 3 5 40.0%
Tracking
10 The virtual hand does not follow my own moves 2 4 30.0%

11 Tool drops from hand 0 3 15.0%
12 Rotating hand does not show well 0 2 10.0%
13 Audio feedback does not differentiate 5 4 45.0%
Feedback
14 Feedback required on gesture rhythm or velocity 4 0 20.0%

15 Low visibility of progress during engraving 2 1 15.0%
16 User stands up (to better position the hand) 3 3 30.0%
User Positioning
17 User places hand at an unusual orientation for better view 2 2 20.0%

18 User keeps both hands in scene to avoid possible tracking loss 2 0 10.0%
19 User places elbow on table to apply gestures 2 0 10.0%
20 The Leap is better positioned on the left hand side of the desk 1 0 5.0%
21 User misses target when it is at the edges of the screen 6 4 50.0%
Control/
reach
22 User would like to pan and position the statue 1 2 15.0%

23 Difficult to select tool (three tools too close) 0 1 5.0%
24 User keeps hand open to apply gestures 6 0 30.0%
Gesturing
25 User attempts to change hand 3 3 30.0%

26 User would like to use both hands (bimanual manipulation) 1 0 5.0%
27 Can we add tattoos? 0 3 15.0%
28 Can we add my name (player)? 0 2 10.0%
Gamification
29 Can we paint the statue? 0 2 10.0%

30 A better background image (sunset) and surroundings 0 1 5.0%
31 Can we add clothes to the statue? 0 1 5.0%
32 Can we add quests (go find the tool in the forest) 0 1 5.0%
33 Funny sound, nice. Can we add more funny sounds? 0 1 5.0%
34 Hand occludes a good view of the digital tool (carving, engraving) 6 0 30.0%
Visibility
35 User would like to zoom in/out the statue 1 0 5.0%

36 During carving marble pieces fall onto parts of the statue 1 0 5.0%
Total 75 77
order to better position their hands in the scene, while others adopted an unusual orien‐
tation of the arm and hand (2/10 adults, 2/10 children) in order to be able to view both
their hand and the digital tool used.
Most interaction design and UX issues are about the look & feel of the application
(24 % of total issues for adults, 38 % for children). In this respect most users (4/10 adults;
9/10 children) reported that they did not have a good sense of depth and they often
performed operations either in front of the statue or behind it. Furthermore, most users
(5/10 adults, 7/10 children) reported arm fatigue especially during the engraving task.
Another issue was that the virtual hand could move through (inside) the statue, which
gave a strange feeling to some users. Other look & feel issues are listed in Table 2.
A considerable number of issues were about feedback (14.7 % for adults; 8.9 % for
children). In this respect, users reported that they would like more elaborate audio feed‐
back on some operations, especially during engraving when the sound of “scratching on
marble” was the same regardless of whether the user was engraving in the correct part
of the statue or not (5/10 adults; 4/10 children). Regarding visual feedback, four adult
user reported that they would need some feedback on whether they have a good appli‐
cation of the gestures in terms or “rhythm” or velocity. Despite that we had provided a
short animated tutorial at the beginning of each task, they would prefer dynamic visual
feedback to adjust their moves.
A number of interesting findings were obtained regarding the user application of
gestures. The most common (6/10 adults) was that users kept their palm open during
some operations, especially during smoothing to enable a better view of the tool used.
We had anticipated that, therefore we had designed the grasping of tools in a form of
having the tool “attached” to the palm during the whole task (we did not want users to
drop tools accidentally or by instant track losses). Notably, this behaviour was not
observed in children interactions, because they did not have tool occlusion issues. In
addition, we observed that some users (3/10 adults; 3/10 children) attempted to grasp
the tool with their left hand, especially to work on the statue from the left side.
One area of findings was about the control and reach of all areas in the scene. We
saw that many users (6/10 adults, 4/10 children) faced difficulties to reach to parts of
the statue when these were at the edges of the scene; users moved their hand outside the
Leap field of view and tracking was temporarily lost. Some user reported that they would
like to hold and pan the statue (1/10 adult; 2/10 children), which would help them over‐
come issues of control and reach.
Another area of findings was about the gamification aspects of the application, which
were reported by children alone. Various issues were proposed by children (like to add
tattoos on the statue, to colour it, etc.), either spontaneously during think-aloud or at the
brief interview at the end of the test.
The last group of findings was about the visibility of various elements of the scene
during user interactions, which were reported only by adults. In this respect we identified
that the virtual hand often occluded a comprehensive view of the tool in hand (6/10
users), which guided users to try several positions of the hand as discussed previously.
One user remarked that an added ability to zoom in/out of the statue would be of
convenience, while another reported that the parts of the statue they wanted to carve
were occluded by previously removed parts (which was accidental for that user and
could be easily removed with either hand).
6 Discussion
We have presented a detailed evaluation of the user experience of the Cycladic sculpture
application which included both formative and summative assessments based on the
concurrent think-aloud protocol, performance measurements and a short questionnaire.
A total of twenty (20) participants tested the application; ten adults and ten children,
who all reported a positive user experience. However particular UX issues remain,
therefore we have identified the following set of guidelines about the design and devel‐
opment of this application, which may be relevant to similar development:
• User Positioning: Ensure that users can apply all required operations with comfort
from the current setup and orientation of the screen, the controller and their stance.
• User Positioning: Ensure that users do not hold their whole hands on the air at a
stretched position for long (not more than 20 s approximately).
• Feedback: Detailed and differentiated audio feedback is required for each (sub) task.
A repertoire of sounds needs to be designed/produced/tested at design phase.
• Feedback: Detailed and differentiated visual feedback is required regarding: (a) the
effect of gestures in the environment, and (b) the rhythm of gesture application.
• Error avoidance: Stick objects in user hand to avoid accidental drops.
• Gesturing: Let the user decide on the manner of holding an object, including the
possibility to change hand.
• Control/reach: Ensure that the scene can be enlarged or scaled down, when the user
applies detailed (or rough) operations.
• Gamification: Embed relevant gamification issues drawn on user requirements.
7 Further Work
The Cycladic sculpture application may be discretely seeded into the physical museum
collections. In contrast to other kinaesthetic installations that are large in scale and
require considerable free space for users to move and room for projected display, this
application may be situated in a kiosk close to the figurines and crafting tools. This would
allow a seamless coupling of physical exhibits and interactive technology in an afford‐
able manner with regard to required space, budget and operational costs.
The application may be conceptually connected to Cycladic figurines that are present
in the physical exhibition. In this way, for example a user might choose to create another
figurine like that of a pregnant woman and in accordance to select appropriate tools and
make respective operations. Furthermore, users could identify these tools or figurines
in the physical collection and be motived to further seek for exhibits; the latter could be
enhanced with mobile technologies and gamification scenarios.
Another area of further work is personalization: users can create some form of
personal identity when they use the app, which can include not only their game scores
but also individual content creations. The latter is related to recommendations made by
many children about how they wanted to change the colours and texture of the figurines,
as well as to make them wear clothes and jewellery. This extension could allow users
to view a gallery of such creations made by previous visitors, as well as to share their
own creations in social media or get a copy of that with 3D printing technology.
A further area of work is to enhance aspects of interactive storytelling features of
this app. According to [11], the state of the art of storytelling in the field of cultural
heritage and virtual museums is still very primitive, with a few good but isolated exam‐
ples. An approach that can be adapted and extended for embodied interactions is outlined
by the CHESS project [13] which offers an infrastructure for authoring cohesive narra‐
tions with carefully-designed references to the exhibits.
References
1. Anderson, E.F., McLoughlin, L., Liarokapis, F., Peters, C., Petridis, P., de Freitas, S.:
Developing serious games for cultural heritage: a state-of-the-art review. Virtual Real. 14(4),
255–275 (2010)
2. Dourish, P.: Where the Action Is: The Foundations of Embodied Interaction. MIT Press,
Cambridge (2004)
3. England, D.: Whole body interaction: an introduction. In: England, D. (ed.) Whole Body
Interaction, pp. 1–5. Springer, London (2011)
4. Fanini, B., d’Annibale, E., Demetrescu, E., Ferdani, D., Pagano, A.: Engaging and shared
gesture-based interaction for museums the case study of K2R international expo in Rome. In:
2015 Digital Heritage, vol. 1, pp. 263–270. IEEE (2015)
5. Hein, G.E.: Learning in the Museum. Routledge, London (2002)
6. Hsu, H.M.J.: The potential of kinect in education. Int. J. Inf. Educ. Technol. 1(5), 365 (2011)
7. Koutsabasis, P., Domouzis, C.: Mid-air browsing and selection in image collections. In:
International Working Conference on Advanced Visual Interfaces (AVI 2016), Bari, Italy, 7–
10 June 2016. ACM (2016)
8. Leap Motion App Store. https://apps.leapmotion.com/
9. Museum of Cycladic Art. http://www.cycladic.gr/
10. Papadimitriou, N.: How Were They Created? Materials and Techniques of Crafting Ancient
Artefacts. A publication of the Museum of Cycladic Art (2015). ISBN 978-618-5060-12-1.
(In Greek)
11. Pietroni, E., Adami, A.: Interacting with virtual reconstructions in museums: the Etruscanning
project. J. Comput. Cult. Heritage 7(2), 9 (2014)
12. Read, J.C.: Validating the fun toolkit: an instrument for measuring children’s opinions of
technology. Cogn. Technol. Work 10(2), 119–128 (2008)
13. Vayanou, M., Katifori, A., et al.: Authoring personalized interactive museum stories. In:
Mitchell, A., Fernández-Vara, C., Thue, D. (eds.) ICIDS 2014. LNCS, vol. 8832, pp. 37–48.
14. Von Hardenberg, C., Bérard, F.: Bare-hand human-computer interaction. In: Proceedings of
the 2001 Workshop on Perceptive User Interfaces, pp. 1–8. ACM, November 2001
15. Vosinakis, S., Koutsabasis, P., Makris, D., Sagia, E.: A kinesthetic approach to digital heritage
using leap motion: the Cycladic sculpture application. In: 8th International Conference on
Games and Virtual Worlds for Serious Applications (VS-GAMES), Barcelona, 5–8
September 2016. IEEE (2016)
16. Vosinakis, S., Xenakis, I.: A virtual world installation in an art exhibition: providing a shared
interaction space for local and remote visitors. In: Rethinking Technology in Museums (2011)
Researching Knowledge Concerns in Virtual Historical
Architecture
S. Münster1 ✉ , C. Kröber1, H. Weller2, and N. Prechtel2

( )
1
Media Center, TU Dresden, 01062 Dresden, Germany
{sander.muenster,cindy.kroeber}@tu-dresden.de
2
Institute for Cartography, TU Dresden, 01062 Dresden, Germany
[email protected]
Abstract. 3D reconstructions have always been an important medium for

teaching, illustrating and researching historical facts and items, especially archi‐
tecture. Virtual representation is often created by cross-disciplinary workgroups,
addressing a wide and heterogeneous audience. The authors investigated knowl‐
edge-related phenomena in four stages, using qualitative and quantitative research
methods. The first stage focuses on the scope and overall relevance of virtual
architecture within the field of digital heritage, and the second investigates
phenomena related to the creation of virtual architectural representations. A third
stage examines how skills and competencies for creating virtual architectural
representations evolve during a project and whether teaching facilitates their
development. Finally, a fourth stage evaluates how to design virtual building
representations to make them comprehensible to a lay audience.
Keywords: Historic architecture · Digital 3D reconstruction · Cultural heritage ·

Visual communication · Knowledge representation · Information sciences
1 Introduction
Starting from the early 1980s, virtual 3D modeling technologies and computer- generated
images have been used frequently in the context of cultural heritage. Related functions rely
primarily on research, education and archiving, as well as the reproduction or management
of sites and items. Architecture and built structures are key objects of both cultural heritage
and digital modeling. In this context, our current research aim is to understand how digital
methods correspond with architectural cultural heritage research and education and how this
relates to the transfer and transformation of knowledge. Particular questions are (a) the
importance of architecture in scholarly discourse on 3D modeling and (b) scenarios,
phenomena and strategies for knowledge transfer and management related to the creation
of virtual 3D reconstructions. Associated interests are identifying the implications for (c)
appropriate teaching of digital 3D modeling methods in higher education as tools for knowl‐
edge communication and (d) designing virtual architectural models to support the percep‐
tion of inherited knowledge of an object.

DOI: 10.1007/978-3-319-48496-9_29
Researching Knowledge Concerns in Virtual Historical Architecture 363
2 Theoretical Concepts
According to Barcelo, we “do not understand past social actions by enumerating [all
possibilities]” [1, p. 414], but need a linkage between digital tools and human interpre‐
tation. Our research focuses on aspects of knowledge transformation and transfer, within
and between humans as well as between humans and—as data—computers. This is
closely related to concepts of intrapersonal knowledge such as reasoning or memoriza‐
tion, including how groups of people communicate and joint mental modeling [2].
Moreover, it is affected by human computer interaction and their explanatory
approaches. Besides a perspective on knowledge holders, concepts like visual reasoning
or embodied knowledge [3, 4] focus on an object which contains and represents knowl‐
edge, for example architecture.
Knowledge-Related Phenomena and Effects of Digital 3D Reconstruction. Why should

we focus on knowledge? The main intention of both research and education is to gain
and transfer knowledge. According to Müller [5], knowledge in the context of visual
media relates to (1) the production of visual media, (2) the visual medium as object and
(3) the reception of visual media. Against this background, digital 3D reconstruction
models and visualizations act as (1) boundary objects—in terms of cross-culturally
understandable media [c.f. 6]—for research and communication in the context of visual
humanities. Furthermore they (2) embody substantial knowledge—in terms of psycho‐
logical-physiological cognition, that is, about the proportions or dimensions of objects
—during the creation of models and visualizations [7]. Last but not least, visual percep‐
tion and reasoning is highly influenced by the properties or Gestalt of visual assets [8],
for example its color or shape. The latter implicitly relates to the level of abstraction [9].
While semiotics focus on a function and Gestalt of signs, model theory focuses on the
relation between an original and a model as its abstraction [10]. An established explan‐
ation schema illustrates Stachowiak’s general model theory. This embodies a simplified
or abridged image of an original, which is “pragmatic” in the sense of its subjective
purpose orientation effect [11].
Digital 3D Reconstruction as a Catalyzer for Knowledge Transfer. As discussed above

and verified empirically [12], digital 3D reconstructions rely on the transfer and incor‐
poration of knowledge from various epistemic cultures. This raises questions about
procedures and strategies for accessing, exchanging, and archiving digital assets along
with the underlying knowledge base. One of the most common explanation models for
cross-disciplinary understanding, the common ground approach, originates in cognitive
psychology and explains factors and recommendations for successful communication
[13]. Successful knowledge transfer between two individuals requires that their knowl‐
edge, beliefs, and assumptions are synchronized [14]. Especially in cross-disciplinary
communication, issues include varying meanings of the same terms across disciplines
or the opposite, usage of different terms for the same content [15]. These difficulties are
exacerbated by the fact that differences are not known in advance, but usually show up
only in the course of interdisciplinary cooperation. While the common ground theory
focuses on communication between individuals, the trading zones approach focuses on
364 S. Münster et al.
modes of information exchange and collaboration in interdisciplinary working teams

[16]. In these contexts, visualization techniques are important for gathering immediate
visual feedback to sharpen and approve design visions. Visual media greatly support the
formation of ideas within research and learning processes [4] and facilitate a deep
learning effect [17].
Digital 3D Reconstruction and Visual Media to Enhance Learning Effects. Another

aspect of interest is the best design for visual media to serve as outcomes of digital 3D
reconstructions. This primarily depends on the type of visual media, whether physical
objects, paper-based depictions, or digital media. Each medium has its own limitations
and benefits: images can be easily remembered [18] and are well-suited to conveying
information about spatial settings and shapes, but, unlike textsare not so suited to
conveying blurry information, such as the various probabilities of reconstruction hypoth‐
eses of historical buildings. Another question concerns degrees of visual freedom,
including qualities of visualization (whether realistic, abstract or schematic), and their
role in knowledge communication and reasoning in certain scenarios [19]. Furthermore,
an important question is how visual media are embedded in multi-sensory settings and
accompanied by interaction—by gestures, pencils, or digital devices—or combined with
auditory or olfactory impressions [c.f. 20].
3 Research Design and Results
Information sciences [21] and visual humanities are the disciplines best equipped to
investigate knowledge-related phenomena associated with visual, virtual representa‐
tions of historical architecture. We combined various theory-driven and empirical
approaches for this research. The leading paradigm was to identify and explore
phenomena and generate hypotheses using qualitative and quantitative research
methods. The analysis was conducted in four stages. The first research stage focused on
the scope and overall relevance of virtual architecture within the field of digital heritage.
The second stage investigated phenomena in creating virtual architectural representa‐
tions, especially the role of visual media. The third stage was a practice project to
examine how cross-disciplinary skills and competencies for creating virtual architectural
representations evolve during a project and whether teaching facilitates their develop‐
ment. A fourth stage evaluated how to design virtual architectural representations to
make them comprehensible to a lay audience.
3.1 Discussing Virtual Representations of Historical Architecture

Research Design. The first step was to identify the scope and overall relevance of virtual
architecture within the scholarly discourse. This included content analysis of publica‐
tions in the field of digital heritage, examining both current projects and related objects.
The data base was 452 journal articles and conference proceedings [12]. Qualitative
content analysis was followed by the development of a categorization and its quantifi‐
cation [22].
Fig. 1. Objects of reconstruction (n = 168, architectural objects highlighted).
Findings and Discussion. The investigated articles mention a total of 168 sites and
items (Fig. 1). Of these objects, 41 are constructed structures, especially religious build‐
ings like shrines, churches or temples, but also residence buildings and sometimes
theater or opera houses (n = 4). A further 21 projects are dedicated to modeling various
types of buildings including villas, city halls or cottages. Another 21 projects model
parts of a city. Numerous projects focusing on modeling other architectural structures,
including building parts like columns and vaults, were subsumed as “others” (n = 8).
The current discourse on digital 3D modeling and reconstruction primarily focuses on
architectural structures. Indeed, “architecture” is the second most important content-
related keyword after cultural heritage used to tag articles in the field of digital recon‐
struction [23]. These findings explain why many attempts at standardization, documen‐
tation or—as shown in the next section—workflows focus specifically on architecture.
3.2 Creating Virtual Representations of Historical Architecture

Research Design. The second stage analysed production scenarios and workflows to
examine knowledge transfer and related phenomena and workflows within project
workgroups and creational processes [24]. This stage contained a survey of 26 project
reports to examine current production scenarios of virtual 3D reconstructions for never
built or no longer extant architecture (Table 1). An investigation took place via inductive
and deductive qualitative content analysis to identify and classify project constellations
and related phenomena. The leading paradigm for an evaluation of this stage was a mixed
methods approach including heuristic frameworks and Grounded Theory [25].
For the four cases, a total of 9 interviews were conducted with key team members.
In addition, 6 direct and participating observations of team meetings were carried out.
A significant number of documents, including log files, communication data, protocols,
sources, and model renderings, were also included in the investigation.
Findings and Discussion. Digital 3D reconstruction (Table 2) creates 3D models from

various sources, such as photographs, plans, and historic drawings. In all observed
projects, especially for architecture that is no longer extant or has never been built,
historic plans and panoramas were by far the most important sources for creating a virtual
reconstruction. A related phenomenon is that digital reconstruction projects require
Table 1. Sample (Projects)

Project Material
Reports 26 project reports describing interpreta‐ 1999–2011
tive 3D reconstruction projects
Cases Roman City (no longer extant) Interviews, documents
Palace Gardens (never realized) Interviews, direct observation, documents
Church Interior (never realized) Interviews, participating observation,
communication data, documents
Necropolis (no longer extant) Published project report, interviews
skills in both highly specialized and complex software and interpretation techniques
common within the humanities. For this reason, most projects are cross-disciplinary
[29].
Table 2. Classification scheme: sources, modeling, and visualization. Based on [26, 27].
Sources Modeling Visualization

Historical sources Semi-automated modeling: Static images
algorithmic reconstruction Animations
Hist. images: panoramas, plans
Procedural generators Interactive
Additional hist. sources: e.g., text
Manual CAD/VR modeling Visualization: i.e.
Contemporary sources games
Data output: e.g., for
Visual: sites, plans,
manufacturing
photography
Data: laser scans,
photogrammetry
Logical sources
Architectural systems
Analogies
Inner model logic
Several factors influence a reconstruction process and lead to certain strategies for
communication, cooperation, and quality management within these projects [28]. Espe‐
cially when working groups with different disciplinary backgrounds are involved, exten‐
sive use of visual media fosters communication and quality negotiations (Fig. 2).
Common quality negotiation strategies deployed in the projects were based on a
comparison between source images and images of the virtual reconstruction. While
widely established symbols like arrows were used in all projects, some projects created
their own graphical codices or metaphors for communication. Moreover, several
projects successfully adopted highly standardized conventions from architectural draw‐
ings for interdisciplinary exchange. Regarding function, the generated 3D models
support the preservation, reconstruction, documentation, research, and promotion of
cultural heritage [37]. The aim is to present the virtual 3D models to an audience
Fig. 2. Exemplified modeling and quality control techniques.
primarily as visual output, which could be static images, animations, or even interactive
visualizations like computer games.
3.3 Learning Virtual Representations of Historical Architecture

Research Design. While the above research stages focused on ongoing or terminated
projects in mostly professional contexts, this third stage investigates personal and organ‐
izational learning processes during a complete project lifecycle, as well as possibilities
to facilitate these processes through education. Despite various attempts and projects to
employ digital 3D reconstruction methods in education, teaching digital 3D reconstruc‐
tion as a skill and as a research and visualization technology still plays just a minor role
in academic instruction. Supported by the Saxon Centre for Teaching and Learning in
Higher Education, an educational project was conducted to create an audio-visual guide
for smartphones which describes certain places of interest and no longer extant construc‐
tion parts of the cathedral in Freiberg in eastern Germany [29]. For the project-based
work, 10 student teams were formed. The humanities students focused on the research
and textual description of certain historical aspects, while geosciences students
performed virtual reconstructions and animations to illustrate the results (Fig. 3). The
course taught and trained relevant competencies, object-related knowledge about the
cathedral, and best practice for project management and communication. The project
was evaluated in several ways. Two joint group discussions—in the middle and at the
end of the project—aimed to identify problems and to estimate the success of didactic
guidance and the students’ adaptation of practices and strategies. In addition, question‐
naires at the beginning of the educational project, at the end of phase one, and at the end
of phase two (n = 15; n = 11; n = 7) queried student satisfaction and their individual
needs. All evaluations were analyzed via inductive and deductive qualitative content
analysis [22].
Findings and Discussion. As a general finding, most of the students involved in the
educational project were highly trained in methods of their own disciplines but had no
previous experience of interdisciplinary teamwork.
The findings of the second stage, mirrored as an implication from our research activ‐
ities, illustrate that the common ground and trading zones essential for cooperation
within virtual architectural representation evolves slowly and mostly due to emerging
Fig. 3. Virtual reconstruction of the cloister
problems. According to cognitive load theory [30, 31], especially the early stages of
project involved a high cognitive workload, affected by the high number of simultaneous
tasks (e.g., team-building, task structuring, and improving the required professional
competencies) [32, 33]. As highlighted in [17], mental processing and problem solving
workflows are very different for textual and visual media. Moreover, the epistemic
cultures [34] of art history and language studies are closer to each other, as humanities
disciplines, than to geosciences. Consequently, synchronization between text and image
related disciplines was especially challenging and time consuming. In problem solving,
solution quality and progress greatly depend on the prior real or imagined experiences
of individuals for a certain situation [35] and the availability of extant patterns for a
solution [36]. This means that teaching implications and best practices prior to a project
is less effective than coaching during it.
Related to the role of architecture within the educational project, several phenomena
can be sketched. The cathedral in Freiberg functioned as a boundary object for commu‐
nication between disciplines, facilitating cross-disciplinary discussion by making design
and spatial relations of architecture directly feasible for all team members. Moreover,
abstract, non-nomothetic information like multiple reconstruction alternatives or rela‐
tional information could be properly transferred via media such as text or speech.
Regarding the resulting application and the ambition to illustrate complex phenomena,
mutual description and depiction enhanced the quality and clarity of information presen‐
tation [37]. The targeted application relied on various media, which required that the
student team members develop competencies for selecting, processing and combining
media, keeping in mind their suitability as a learning tool for visitors to the cathedral.
Encouraging students to research and explore historical objects by reproducing them
virtually is an increasingly important issue [38]. Previous research has already focused on
the role and effects of media in higher and museum education, especially visualizations as
outcomes of digital 3D reconstruction processes [e.g. 39–41]. Other researchers have inves‐
tigated the design of learning scenarios to educate about digital cultural heritage objects [e.g.
42, 43]. Sahle [44] has defined a core curriculum which is intended to serve as a blueprint
for a design of academic digital humanities courses. While Liu [45] described general
requirements for scholars in the field of humanities, Sprünker [46] defined a set of compe‐
tencies required to cope specifically with digital 3D reconstruction and visualization.
Despite all these activities, there is still no broad consensus on a specific digital visual
humanities education paradigm, and there is still a lack of larger studies on teaching digital
methods in visual humanities [c.f. 47, p. 405].
3.4 Designing Virtual Representations of Historical Architecture

Research Design. A fourth stage investigated the practical implications and user testing
of a virtual architectural design. While game studies focus on final products [48] and
their relation to an original, our interest was to explore how various levels of detail would
affect the recognition of buildings and city structures. We developed six varying designs
for a no longer extant block of buildings, which was formerly located in the city of
Dresden in Germany and originally built in the late 19th century. The evaluation included
two studies to (a) compare design alternatives for the buildings, and (b) verify the visual
usability of the formerly preferred design alternative. Both studies employed a thinking
aloud methodology from usability testing [49, 50], in which test users perform several
given tasks within an application and comment on their activities and thoughts. In the
first case 6 users (aged 23–52, 2 male and 4 female) compared and negotiated 6 design
alternatives [51]. A second test involved 15 people with various levels of expertise
concerning virtual architectural representation (5 experts, 4 novices, 6 no answer). This
group performed tasks including identifying certain buildings within a virtual city model
and describing their unique features, such as the type and certain attributes of the
building.
Results and Discussion. A well-established approach to categorizing qualities of archi‐

tectural representation is the level of detail (LOD) scheme. LOD distinguishes between
five different levels: LOD 1 is a block model or extruded footprint of a building and
LOD 2 includes the block, its height and roof design. LOD 3 adds a photo texture of a
facade, while LOD 4 stands for highly detailed modeling of the building hull, including
all unique features. Finally, LOD 5 is highly detailed modeling of inner and outer
building features [52]. While the LOD scheme provides just a rough classification, it
may help define which features are most essential to identify a single building within a
block and to identify a building as unique in comparison to others. The research literature
suggests the following potentially influential features:
• Basic shape, including a simplified footprint of the building;
• Building height;
• Roof shape;
• Façade photo texture (e.g., windows, storeys);
• Building details (e.g., chimneys, balconies).
Expanding on this classification, six design alternatives with varying features were
created (Fig. 4). All tests were performed within a modified Google Earth environment,
including several degrees of freedom for the user to change the viewpoint. More gener‐
ally, identifying a building and estimating its scale in a virtual environment is highly
dependent on the (virtual) context and any given information [53]. Especially block
models (LOD 1) are insufficient by themselves. The additional information needed can
be provided either by a higher level of detail in the case of a single building (e.g., by
adding roof structures), by providing auxiliary visual clues (e.g., a map overlay as a
ground plan for a building), or by combining various buildings with differing heights
into blocks. Concerning the identification of individual buildings within a block, espe‐
cially the basic shape, unique height, and individual roof colour foster recognition. An
individual façade photo texture is perceptible only in close views. The basic geometry,
height and individual roof shape of a building provide important visual clues to identify
it as unique. The façade and additional building details influence this decision to a lesser
extent. To sum up, little visual information is needed to identify a single building and
to distinguish buildings from each other, (relatively) regardless of whether the users are
novices or experts. How level of expertise and a professional background affect recog‐
nition of architecture is an interesting subject for future research.
Fig. 4. Design alternative example
4 Conclusion
The architecture of constructed structures is the most relevant object category for 3D
modeling in cultural heritage and most prominent in academic discourse. Especially in
the humanities, scholars do not yet have much competence in or affinity for digital
research methods [56]. The key challenge is still to develop the skills of scholars and
users in producing, evaluating and using digital reconstructions.
Creating a common ground for understanding is another major challenge in cross-
disciplinary projects. Language is a sophisticated code pattern and requires the synchro‐
nization of individual interpretation and understanding. Architectural structures and
their visual representation are highly suitable for creating common ground for interdis‐
ciplinary and expert-novice communication. They function as a boundary objects, simi‐
larly to shared metaphors. Especially for communication and negotiation, visual media
are widely used within a creation workflow. Related competencies draw far more on
implicit knowledge and experience than on theoretical implications and best practice.
While project work, cross-disciplinary cooperation, and complex problem solving
processes are an essential part of modern professional life and visualization projects,
academic learning mostly focuses on explicit, theoretical knowledge and disciplinary
ambition, rather than its practical application in complex scenarios. Paying more atten‐
tion to project-based learning and related competencies in academic education in general
would likely enhance cooperation quality and project progress.
Regarding the level of detail required to recognize architectural buildings, relatively
little visual information is needed to allow viewers to distinguish buildings from each
other or to identify a single building and to gain information about its spatial relation
and shape. Therefore, architectural and landscape models provide an appropriate
approach to structure and access further space-related information in a visual way,
including historical sources, background information, and relationships [54]. As demon‐
strated in our research, even high levels of visual abstraction and schematic appearance
can make architecture recognizable to an observer. This makes it possible to visualize
different source qualities and levels of certainty, for example by schematic visualization
[55–57]. This explorative study relies on methods from the social sciences. Various
concepts and phenomena can be sketched, but the data available only allows for limited
more general implications.
Acknowledgements. The Freiberger Dom educational project was funded by the Saxon Center
for Higher Education in 2014 within the Learning in Transfer scheme. Research on Usability
aspects was supported by Josefine Brödner and Katharina Hammel. The research activity
described in this paper was funded by the German Federal Ministry of Education and Research
(no. 01UG1520).
References
1. Barceló, J.A.: Towards a True Automatic Archaeology: Integrating Technique and Theory.
Universitätsbibliothek der Universität Heidelberg, Heidelberg (2010)
2. Cannon-Bowers, J.A., Salas, E., Converse, S.A.: Shared mental models in expert team
decision making. In: Castellan Jr., N.J. (ed.) Individual and Group Decision Making: Current
Issues, pp. 221–246. Lawrence Erlbaum Associates Inc., Hillsdale (1993)
3. Arnheim, R.: Visual Thinking. Rütten & Loening, München (1969)
4. Gooding, D.C.: Cognition, construction and culture. Visual theories in the sciences. J. Cogn.
Cult. 4, 551–593 (2004)
5. Müller, M.G., Geise, S.: Grundlagen der visuellen Kommunikation: Theorieansätze und
Analysemethoden. UVK Verlagsgesellschaft, Konstanz (2015)
6. Star, S.L., Griesemer, J.R.: Institutional ecology, ‘Translations’ and boundary objects.
Amateurs and professionals in Berkeley’s Museum of Vertebrate Zoology 1907–1939. Soc.
Stud. Sci. 19(4), 387–420 (1989)
7. Wilson, R.A., Foglia, L.: Embodied cognition. In: Stanford Encyclopedia of Philosophy,
Stanford (2015)
8. Gerrig, R.J., Zimbardo, P.G.: Psychology and Life, 19th edn. Allyn & Bacon, Boston (2010)
9. Tversky, B.: Visuospatial reasoning. In: Holyoak, K., Morrison, R. (eds.) Handbook of
Reasoning, pp. 209–249. Cambridge University Press, Cambridge (2005)
10. Mahr, B.: Das Wissen im Modell. Technische Universität, Fakultät IV, Berlin (2004)
11. Stachowiak, H.: Allgemeine Modelltheorie. Springer, Wien (1973)
12. Münster, S., Köhler, T., Hoppe, S.: 3D modeling technologies as tools for the reconstruction
and visualization of historic items in humanities. A literature-based survey. In: Traviglia, A.
(ed.) Across Space and Time, Papers from the 41st Conference on Computer Applications
and Quantitative Methods in Archaeology, Perth, 25–28 March 2013, pp. 430–441.
Amsterdam University Press, Amsterdam (2015)
13. Repko, A.: Integrating interdisciplinarity. How the theories of common ground and cognitive
interdisciplinarity are informing the debate on interdisciplinary integration. Issues Integr.
Stud. 25, 1–31 (2007)
14. Clark, H.H.: Using Language. Cambridge University Press, Cambridge (1996)
15. Bromme, R.: Beyond one’s own perspective. The psychology of cognitive interdisciplinarity.
In: Weingart, P., Stehr, N. (eds.) Practising Interdisciplinarity, pp. 115–133. University of
Toronto Press, Toronto (2000)
16. Galison, P.: Image & Logic: A Material Culture of Microphysics. University of Chicago Press,
Chicago (1997)
17. Mintzberg, H., Westley, F.: Decision Making: It’s Not What You Think. In: Nutt, P.C.,
Wilson, D. (eds.) Handbook of Decision Making, pp. 73–82. Wiley-Blackwell, Oxford (2010)
18. Nelson, D.L.: Remembering pictures and words. Appearance, significance, and name. In:
Cermak, L.S., Craik, F.I.M. (eds.) Levels of Processing in Human Memory, Mahwah, pp. 45–
76 (1979)
19. Bertin, J., Jensch, G.: Graphische Semiologie: Diagramme, Netze, Karten. de Gruyter, Berlin
(1974)
20. Ch’ng, E.: Experiential archaeology: Is virtual time travel possible? J. Cult. Heritage 10(4),
458–470 (2009). doi:10.1016/j.culher.2009.02.001
21. Stock, W., Stock, M.: Handbook of Information Science. De Gruyter Saur, Berlin/Boston
(2015)
22. Mayring, P.: Qualitative content analysis. Forum Qualitative Sozialforschung 1(2), Art. 20
(2000)
23. Münster, S., Kröber, C., Weller, H., Prechtel, N.: Virtual reconstructions of historical
architecture as media for visual knowledge representation. In: Ioannides, M., Magnenat-
Thalmann, N., Papagiannakis, G. (eds.) Mixed Reality and Gamification for Cultural Heritage.
LNCS. Springer, Cham (in print)
24. Münster, S.: The role of images for a virtual 3D reconstruction of historic artifacts. In:
International Communication Association (ICA) Annual Meeting, 17–21 June 2013, London
(2013)
25. Bryant, A., Charmaz, K.: The SAGE Handbook of Grounded Theory. SAGE, Thousand Oaks
(2010)
26. De Francesco, G., D’Andrea, A.: Standards and guidelines for quality digital cultural three-
dimensional content creation. In: Ioannides, M., Addison, A., Georgopoulos, A., Kalisperis,
L. (eds.) Digital Heritage: Proceedings of the 14th International Conference on Virtual
Systems and Multimedia, Project Papers, pp. 229–233. Archaeolingua, Budapest (2008)
27. Hermon, S.: Reasoning in 3D. A critical appraisal of the role of 3D modelling and virtual
reconstructions in archaeology. In: Frischer, B. (ed.) Beyond Illustration: 2D and 3D Digital
Technologies as Tools for Discovery in Archaeology, vol. 1805, pp. 36–45. Tempus
Reparatum, Oxford (2008)
28. Münster, S.: Workflows and the role of images for a virtual 3D reconstruction of no longer
extant historic objects. In: ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial
Information Sciences (XXIV International CIPA Symposium), vol. II-5/W1, pp. 197–202
(2013)
29. Kröber, C., Münster, S.: Educational App creation for the Cathedral in Freiberg. In: Spector,
J.M., Ifenthaler, D., Sampson, D.G., Isaias, P. (eds.) Competencies, Challenges, and Changes
in Teaching Learning and Educational Leadership in the Digital Age. Springer, Switzerland
(2016)
30. Collins, A., Brown, J.S., Newman, S.E.: Cognitive apprenticeship: teaching the craft of
reading, writing and mathematics (Technical Report No. 403). University of Illinois,
Cambridge (1987)
31. Hatcher, T., Hinton, B., Swartz, J.: Graduate student’s perceptions of university team-
teaching. Coll. Student J. 30(3), 367–376 (1996)
32. Wegner, D.M.: Transactive memory. A contemporary analysis of the group mind. In: Mullen,
B., Goethals, G.R. (eds.) Theories of Group Behavior, pp. 185–208. Springer, New York
(1986)
33. Argyris, C., Schön, D.A.: Organizational Learning: A Theory of Action Perspective. Addison-
Wesley, Reading (1978)
34. Knorr-Cetina, K.: Epistemic Cultures: How the Sciences make Knowledge. Harvard
University Press, Cambridge (1999)
35. Tsoukas, H.: Complex Knowledge. Oxford University Press, Oxford (2006)
36. Simon, H.A.: Theories of bounded rationality. In: McGuire, C.B., Radner, R. (eds.) Decision
and Organization, pp. 161–176. North-Holland Publishing Company, Amsterdam (1972)
37. Paivio, A.: Dual coding theory and education (Draft). In: Pathways to Literacy Achievement
for High Poverty Children. The University of Michigan School of Education, 29 September–
1 October 2006
38. Gerth, B., Berndt, R., Havemann, S., Fellner, D.W.: 3D modeling for non-expert users with
the castle construction kit v0.5. In: Mudge, M., Ryan, N., Scopigno, R. (eds.) 6th International
Symposium on Virtual Reality, Archaeology and Cultural Heritage (VAST 2005), pp. 49–57.
Eurographics Association, Pisa (2005)
39. Sanders, D.H.: Virtual archaeology: yesterday, today, and tomorrow. In: Niccolucci, F.,
Hermon, S. (eds.) CAA 2004, Prato, 13–17 April 2004. p n.a.
40. Fisher, C.R., Terras, M., Warwick, C.: Integrating new technologies into established systems:
a case study from Roman Silchester. In: Computer Applications to Archaeology 2009,
Williamsburg, Virginia, USA, 22–26 March 2009
41. Flaten, A.: Ashes2Art: a pedagogical case study in digital humanities. In: CAA (2008)
42. Ferrara, V., Macchia, A., Sapia, S.: Reusing cultural heritage digital resources in teaching. In:
Digital Heritage International Congress (DigitalHeritage), 28 October–1 November 2013, pp.
409–412 (2013). doi:10.1109/DigitalHeritage.2013.6744792
43. Gicquel, P.Y., Lenne, D., Moulin, C.: Design and use of CALM: an ubiquitous environment
for mobile learning during Museum visit. In: Digital Heritage International Congress
(DigitalHeritage), 28 October–1 November 2013, pp. 645–652 (2013). doi:10.1109/
DigitalHeritage.2013.6744831
44. Sahle, P.: Auf dem Weg zu einem Kern- und Referenzcurriculum der Digital Humanities.
Georg-August-Universität Göttingen, Göttingen (2013)
45. Liu, A.: Digital humanities and academic change. Engl. Lang. Notes 47, 17–35 (2009). Special
Issue on Experimental Language Education
46. Sprünker, J.: Educació patrimonial mitjançant recursos educatius en línia amb contingut de
patrimoni cultural i xarxes d’aprenentatge, Ph.D. thesis, Barcelona (2011)
47. Sprünker, J.: Making on-line cultural heritage visible for educational proposes. In: Digital
Heritage International Congress (DigitalHeritage), 28 October–1 November 2013, pp. 405–
408 (2013). doi:10.1109/DigitalHeritage.2013.6744791
48. Torn, J.L.: The virtual cityscapes of rock star games. In: International Communication
Association (ICA) Annual Meeting, London, 17–21 June 2013
49. Krug, S.: Don’t Make Me Think! A Common Sense Approach to Web Usability, 2nd edn.
New Riders Pub, Berkeley (2006)
50. Rubin, J., Chisnell, D.: Handbook of Usability TestingTesting, Second Edition: How to Plan,
Design, and Conduct. Wiley Publishing Inc., Indianapolis (2008)
51. Weller, H.: Generalisierte 3D-Gebäuderepräsentation im Spannungsfeld von
Primärinformation, Modellierungsaufwand und Wiedererkennbarkeit am Beispiel eines 3D-
Stadtmodells von Dresdens um 1940 (Diploma thesis). TU Dresden, Dresden (2013)
52. Pomaska, G.: Web-Visualisierung mit Open Source - Vom CAD-Modell zur Real-Time-
Animation. Herbert Wichmann Verlag, Heidelberg (2007)
53. Yaneva, A.: Scaling up and down. extraction trials in architectural design. Soc. Stud. Sci.
35(6), 867–894 (2005). doi:10.1177/0306312705053053
54. Prechtel, N., Münster, S.: Cultural heritage in a spatial context – towards an integrative,
interoperable, and participatory data and information management. In: Münster, S., Pfarr-
Harfst, M., Kuroczyński, P., Ioannides, M. (eds.) How to Manage Data and Knowledge
Related to Interpretative Digital 3D Reconstructions of Cultural Heritage? Springer, Cham
(in print)
55. Danielová, M.: Visual reconstruction of archaeological data of the Sanctuary of Diana at
Nemi, Italy (Master thesis), Munich (2014)
56. Lengyel, D., Toulouse, C.: Darstellung von unscharfem Wissen in der Rekonstruktion
historischer Bauten. In: Heine, K., Rheidt, K., Henze, F., Riedel, A. (eds.) Von Handaufmaß
bis High Tech III. 3D in der historischen Bauforschung. Verlag Philipp von Zabern,
Darmstadt, pp. 182–186 (2011)
57. Apollonio, F.I., Gaiani, M., Sun, Z.: 3D modeling and data enrichment in digital reconstruction
of architectural heritage. In: International Archives of the Photogrammetry, Remote Sensing
and Spatial Information Sciences, XXIV International CIPA Symposium, 2–6 September
2013, Strasbourg, vol. XL-5/W2, pp. 43–48 (2013)
Project Paper: Data Acquisition, Process
and Management in Cultural Heritage
Surveying Illusory Architectures Painted
on Vaulted Surfaces
Matteo Flavio Mancini1 and Marta Salvatore2 ✉

( )
1
Department of Architecture, Roma Tre University, Rome, Italy
[email protected]
2
Department of History, Representation and Restoration of Architecture,
Sapienza University of Rome, Rome, Italy
[email protected]
Abstract. This paper addresses the problem of surveying illusory architectures

painted on vaulted surfaces. The survey of a quadratura, or a painting in general,
requires recording the metric and chromatic characteristics of the subject and the
typical characteristics of the painted surfaces, such as soot, engravings, and gior‐
nate. Our goals are the proposal of an “optimum” quality standard for surveying
curved painted surfaces and testing a method to acquire and render the data that
allows those standards to be met. The test, conducted on the corridor of Saint
Ignatius of Loyola rooms in Rome depicted by Andrea Pozzo, shows how the
quality of the texture can be measured in terms of overall sharpness and average
resolution. It is also shown how it is possible to identify some reference standards
that allow the quality of the final result to be determined already in the photog‐
raphy phase of the project.
Keywords: Architectural perspective · Photogrammetric survey · Resolution ·

Painted vault · Andrea Pozzo
1 Introduction
This paper addresses the problem of surveying illusory architectures painted on vaulted
surfaces. As is known, illusory architectures, or quadraturas, are perspectives in a
natural scale painted on the walls and ceiling of a room. They represent architectural
subjects painted in continuity with the surrounding real space, thus altering the depth of
the space to the eyes of the observer.
The survey of a quadratura, or a painting in general, requires recording the metric
and chromatic characteristics of the subject and the typical characteristics of the painted
surfaces, such as soot, engravings, and giornate, which are often invisible to the human
eyes unless the artwork is viewed from very close.
The planar or curved morphology of the painted surface determines the survey
method and the type of information it is possible to obtain.
In the case of paintings on flat surfaces, great results can be obtained by constructing
High-Resolution Images (HRI) with metric value [1]. These images, which create a high-
resolution reproduction in proper form of the wall, also contain metric and chromatic

DOI: 10.1007/978-3-319-48496-9_30
378 M.F. Mancini and M. Salvatore
information and allow the painting to be explored immersively, as if it were being

observed up close; the closer it is observed, the higher the resolution.
In the case of vaulted subjects, however, the curvature of the surface requires the
painting to be reconstruct in three-dimensional space. Although it is necessary to guar‐
antee a close reading of the traces of the painting, HRI cannot provide any metric infor‐
mation in this case.
The experiments, which are still in progress1, aim to highlight the main problems
related to acquiring painted vaults. The objectives are the following:
– identify an “optimum” quality standard for surveying curved painted surfaces;
– test a method to acquire and render the data that allows those standards to be met.
2 Image-based 3D Models for Surveying Illusory Architectures
In the study of illusory architectures, it is generally necessary to be able to completely

or partially redrawing the traces of the paintings. This requirement arises from different
needs, which range from the study of the relationship between the illusory space of the
perspective and real space to the study of hypotheses about the traces left in the painting
when transferring the sketches onto the vault; to the construction of the flat developments
of the painted surfaces; to the construction of thematic maps of the painting that highlight
specific types of signs.
To carry out these steps, it is necessary to have a metrically accurate three-dimen‐
sional survey model that can render the proper form of the painting and the vaulted
surface on which it lies without altering the chromatic characteristics. As well, a high
overall resolution should be achieved to make the constructive signs of the painting
legible. Finally, a model of this type should be manageable with common NURBS
representation software in order to construct thematic vector maps.
Having to survey painted surfaces, we focused on techniques capable of faithfully
revealing the chromatic characteristics of the surface. In particular, we used image-based
3D modelling photogrammetry techniques2 aimed at constructing texturized three-
dimensional models whose metric accuracy is entrusted to the images themselves [2–4].
The continual development of computational algorithms used by image-based
software leads to increasing metric accuracy in building the discrete point cloud model.
The correspondence between points in the cloud and the relative chromatic information
is ensured by the fact that both are defined starting from the same set of data: the photo‐
graphs3.
1
We are currently testing this methodology on similar paintings, such as on the vault of the
Church of St. Ignatius and that on the hall of Palazzo Barberini (Rome), to confirm its validity.
2
In this experimentation we chose to adopt one of the currently most popular photogrammetric
software, Agisoft Photoscan 1.1, but the considerations of this paper are still valid by using
any other similar software.
3
The procedure is also advantageous if viewed in light of the time and cost entailed in the
photography and processing phases, which make this technology a shared standard for quick
survey campaigns.
Surveying Illusory Architectures Painted on Vaulted Surfaces 379
Chromatic quality instead depends on the quality of the texture. Since the resolution
and texture definition depend on those in the snapshots, it is currently possible to obtain
a decidedly higher quality with this technique than with those obtainable with the
cameras generally included with the most common laser scanners4.
The overall texture quality is the main necessary characteristic of the survey models
of the painted surface, which are destined for study, critical interpretation, and vector
reconstruction of the illusory architectures. It is determined by the different parameters
that intervene in the different steps to construct the model: acquisition, data processing,
and vectorizing the traces.
For each of these steps, the main parameters that contribute to defining the overall
quality of the survey model and its rendering were identified.
The experimentation was made on the quadraturas painted by Fr. Andrea Pozzo
between 1681 and 1686 in the corridor that leads to the Saint Ignatius of Loyola rooms
in the Church of the Gesù in Rome. This is a favourite case study because it is configured
as a small immersive space in which the observer is completely enveloped by the illusory
space, painted in part on the walls and in part on the vault (Fig. 1).
Fig. 1. Andrea Pozzo, corridor of the Saint Ignatius of Loyola rooms (Rome, 1681–1686).
4
The resolution of cameras installed on today’s commercially available scanners guarantee a
maximum of 4 megapixels on average, compared to the 18 megapixels available on average
for a full-frame professional camera.
3 Quality Parameters in the Photography Phase
In addition to its metric accuracy, the overall quality of an image-based survey is influ‐
enced by the photographic quality of the snapshots, which in turn depends on different
factors including the sharpness and resolution.
If the overall sharpness is measured as a function of different parameters such as the
resolution, acutance, and contrast, the resolution is highly dependent on the dimensions
of the camera sensor. This is expressed as the common pixel/length ratio, and contributes
to defining the quantity of informational content in an image, and therefore its level of
detail [5].
The resolution of a snapshot is also influenced by the size of the sensor, the distance
from the sensor to the subject, and the focal length. These quantities are related by the
following simple relationship:
ls : f = x : d,
where ls is the length of one side of the sensor, f is the focal length, d is the object
distance, and x is the length of the portion of wall in the picture (Fig. 2).
Fig. 2. The calculation of shooting resolution.
It is therefore possible to establish a priori the average resolution of a snapshot, which

determines the average resolution of the model’s texture, calculated based on the project
to acquire the individual snapshots.
A decay, which is due to the curved morphology of the support and is difficult to
quantify, should be considered when calculating the resolution. In fact, the sensor pixels
record a variable unit of length as a function of the orientation of the plane tangent to the
surface, which is different at each point with respect to the plane of the camera sensor in
the case of vaulted surfaces; the higher the curvature of the surface, the greater the decay.
The corridor is shaped as a trapezoid covered by a segmental barrel vault. Its dimen‐
sions are similar to a prism of about 18 × 4.5 m, 3.5 m high imposts and 4.8 m high at
the keystone. Dealing only with the painted surfaces, 174 snapshots were taken with a
Nikon D800E, full-frame sensor (35.9 × 24 mm, 7360 × 4912 pixel, 36.3 megapixel),
with fixed optics and a 28 mm focal length (AF-S NIKKOR 28 mm F/1.8 G).
Initially establishing an average distance of about 2 m from the painting, each indi‐
vidual snapshot recorded a 2.56-m-long portion of wall, with an average resolution of
72 dpi in 1:1 scale, so 30 pixels describe about 1 cm along the wall. This resolution is
adequate for a video exploration of the painting, which is particularly effective in
rendering the “surface” characteristics of the painting such as soot, engravings, and
giornate, and which, operating in these dimensions, does not require particularly heavy
computations.
The intrinsic parameters of the camera also contribute to the overall quality of a
snapshot. These range from the use of low ISO to setting the aperture at a medium-high
value to ensure great depth of field without introducing the diffraction typical of small
apertures, the constant setting of white balance for the entire set of snapshots. The
lighting conditions of the environment also greatly affect the quality of the model, and
the texture in particular. The homogeneity of lighting in the environment is an important
parameter because it guarantees the correct recognition of homologous points when
processing the data and it permits construction of a relaxed surface mesh (characterized
by a low level of noise) and the subsequent mosaicking of a homogeneous texture.
Inhomogeneous exposure for the snapshots can instead cause errors when the software,
which sets its algorithms based on the chromatic values of individual pixels, interprets
the geometry. The rooms of the quadraturas have low lighting or, as in the case of the
corridor of the Church of the Gesù, are lit by openings placed along a single wall, which
light the environment inhomogeneously and make one of the walls strongly backlit.
Homogeneous lighting conditions can be obtained with artificial lighting systems,
for example, inflatable lighting apparatus or flash systems [6] arranged in a fixed position
in the rooms, or, by using high dynamic range (HDR) techniques to compensate zones
with different lighting.
The cinematics of the photography also affect the quality of the model and the related
texture. When we take a photo of a painting, the resulting image is a perspective. If we
imagine taking two snapshots of the same subject, one very close and the other orthog‐
onal to the wall, they will be rendered with a different number of pixels. One will there‐
fore contain richer informational content than the other. To avoid this situation, it is
useful to take snapshots such that the camera’s sensor is parallel to the surface, thus
guaranteeing greater execution of the structure from motion (SFM) algorithms and also
a homogeneous quality of texture detail.
4 Quality Parameters for Data Processing
The photogrammetry software programs use SFM algorithms to reconstruct the three-
dimensional coordinates of points starting with a set of structured images. The pipeline
runs through the same steps in all software: camera calibration, internal and absolute
orientation of the snapshots, reconstruction of geometries, and texture mapping5.
5
The calculations made by the software are statistical, which leads to variations of the results
even when starting with identical datasets.
The geometric quality of the model depends on different parameters, which range
from the snapshot quality to the development settings of the photographic set, to the
computational capacity of the hardware used. Possible compressions of the source files
made when converting the RAW format also influence the geometric quality of the
model, but it is difficult to establish the amount.
The basic quality of the model is entrusted to the calibration and orientation phase.
Making use of the computational characteristics required by the software, it is preferable
to work with the highest accuracy possible. The resolution of the collinearity equations
to recognize homologous points among the snapshots forms the basis of the calibration
and orientation processes. The identification of notable points is made by analyzing the
chromatic gradients around the points and this leads to a high sensitivity to strong-
contrast situations. This is the case of shadows produced by strong natural or artificial
lighting, but also by particularly contrasting chiaroscuro effects present in some quad‐
raturas, as in the vault in the corridor of the Church of the Gesù, where the drawing on
part of the vault and the corners of the beams produce a light relief in the 3D model that
does not exist in reality [7] (Fig. 3).
Fig. 3. The light relief due to the contrasting chiaroscuro effects of the painting.
Texture creation constitutes the last phase of the process to generate the model.
Photogrammetry software programs allow the UV mapping of the surface to be made,
which is used to establish the resolution and method of fusion. In the case of the quad‐
raturas, optimal results are obtained by using fusion methods that consider the normal
to the texturized face. The information is used to choose the best snapshot to extract the
chromatic data, that is, the snapshot made with the camera sensor as parallel to the
surface as possible.
In the case of surveying illusory architectures or paintings in general, the texture is
entrusted with rendering the entire informational content desired. The quality of the
texture is measured mainly in terms of resolution, that is, the level of detail contained,
which is configured as the main quality parameter to establish.
As mentioned above, a set of snapshots has an average resolution. This resolution,

which we call acquisition resolution, decays more or less notably during data processing,
even in relation to the dimensions of the set of snapshots.
Photogrammetry programs construct UV mappings of the acquired surface, which
can be used to establish the dimensions compatible with the computational resources of
the hardware. In the case of the corridor of the Church of the Gesù, the painted surface
covers about 240 m2, equal to a square of about 15 m on a side. It was acquired at
72 pixels/inch, which means that each cm2 of surface was recorded on 900 pixels. If we
wanted to construct a UV texture without losing quality and if it were possible to generate
it by completely saturating the raster space available, the software would have to
construct a map equal to 45000 × 45000 pixels.
If we also consider that a UV map cannot cover the entire square surface used6, we
would have to increase the number of required pixels, achieving dimensions that are
even more onerous for common hardware to calculate.
To resolve this problem, photogrammetry programs allow the texture to be divided
into multiple parts. Therefore, dividing a single texture into smaller maps, it would
theoretically be possible to preserve the acquisition resolution. If we stay in the photo‐
grammetry environment, the question could also be considered solved. Instead, having
to export the model into a software environment dedicated to three-dimensional model‐
ling, we would have to face the maximum resolution of the UV texture managed as a
function of the software and the graphics cards used.
This resolution is generally limited to 8192 × 8192 pixels. Experimental tests have
confirmed this limit; in fact, above these values, NURBS representation programs such
as Rhinoceros, which was used in this experiment, generate a quadrangular tessellation
of the texture that notably reduces the quality.
Therefore, in order to deal with these limits and maintain quality with respect to
the acquisition resolution, in the case of the corridor we should required the photo‐
grammetry software to construct about thirty textures. This is a heavy calculation that
does, nevertheless, act as a useful reference to evaluate the quality. In fact, by
reducing the number of textures, the resolution is also reduced proportionally, which
can be calculated and evaluated qualitatively at this point using the acquisition reso‐
lution as the ideal reference.
The experimentation led to the construction of a model of the corridor of the Church
of the Gesù with an accuracy equal to 8 mm, mapped with 10 textures with a resolution
equal to one-third of the acquiring resolution (Fig. 4).
6
A UV texture associates a raster image to each triangle of a mesh surface developed on the
plane. Since the meshes are irregular polyhedrons, the constituent polygons tend to overlap in
plane development. To avoid this superposition in the development phase, the surface is
divided into multiple parts, and the developed polygons are situated in order to optimize the
raster space available. The developed mesh obviously does not cover the entire square surface
of the texture.
Fig. 4. Comparison between the acquisition resolution (left) and the 3D model resolution (right);
loss of detail of the model texture with respect to the acquired resolution (centre).
5 Vector Reconstruction of Illusory Architectures
The texturized polygonal model serves as the survey basis to initiate the vector redesign
of the traces of the painting on the mesh surface.
The geometric quality of the redrawing operation depends on the quality of the
survey mesh, whose accuracy depends instead on the resolution and sharpness of the
texture since this affects the possibility of identifying a greater number of signs.
Rhinoceros, the software used in this experimentation, includes two design tools
capable of constructing both polylines and curves pertaining to a mesh surface. The
procedure to construct the geometric entities is based on the projection of the entity
represented above the reference mesh projected perpendicular to the plane of the
computer screen. This procedure makes the redrawing operation rigorous. In fact, when
we observe a three-dimensional texturized model, what we see is precisely an orthogonal
projection of the texture drawing on the surface of the monitor. The use of a curve
guarantees maximum adherence to the model because it requires continual input to select
the entire chosen profile, but it also requires particular skill in execution. The use of a
polyline instead provides a redrawing similar to what could be done years ago with an
analogue photogrammetry instrument because the reconstruction of traces was made by
selecting notable points on the surface in order to set the path of straight line segments.
The intermediate segments are interpolated, which introduces a slight approximation,
but the overall control is more accurate.
With these tools it is possible to obtain vector models of the painting or its parts,
which lend themselves to different uses and which vary according to the texture reso‐
lution, like, for example, combination models of the painting or thematic maps that
record particular signs present on the painted surface, such as soot, engravings, and
giornate (Fig. 5). The vector model is also susceptible to reprojection operations and
development useful to study the techniques of transferring the sketch to the painted
surface; it is configured as a useful resource for informative applications.
Fig. 5. The 3D vector reconstruction of the corridor with the engravings of graticola (red lines).
6 Conclusions
The experimentation, which is still underway, aimed to investigate the problem of

surveying painted subjects, with particular reference to illusory architectures. The goal
was to address the main problems related to the acquisition of this type of artwork and
to test an acquisition and reconstruction method capable of producing metrically reliable
three-dimensional models from which thematic vector models can be obtained.
Particular attention was dedicated to assessing the overall quality of these survey
models, which depends on the quality of the surface geometry and the quality of the
texture with which the model is mapped. It was shown how the quality of the texture
can be measured in terms of overall sharpness and average resolution. It was also shown
how, based on these parameters, it is possible to identify some reference standards that
allow the quality of the final result to be determined already in the photography phase
of the project, and therefore to plan the level of detail based on the aims of the survey.
Moreover the validity of this survey method is confirmed by the experimentation
carried out on the corridor of rooms of Saint Ignatius where it leads to the identification
of the engravings needed to apply the construction of the graticola (grid). This construc‐
tion, well known for its application on the Glory of Saint Ignatius in the Church of Saint
Ignatius in Rome, was also used here by Andrea Pozzo to draw the rough sketch on the
surface of the vault. The studies on this topic are currently in progress.
References
1. Migliari, R., Fasolo, M., Baglioni, L., Salvatore, M., Romor, J., Mancini, M.F.: Architectural
perspectives survey. In: Ioannides, M., Magnenat-Thalmann, N., Fink, E., Žarnić, R., Yen, A.-Y.,
Quak, E. (eds.) EuroMed 2014. LNCS, vol. 8740, pp. 463–472. Springer, Heidelberg (2014). doi:
10.1007/978-3-319-13695-0_45
2. Guidi, G., Remondino, F.: 3D modelling from real data. In: Alexandru, C. (ed.) Modelling and
Simulation in Engineering. InTech (2012)
3. Luhmann, T., Robson, S., Kyle, S., Harley, I.: Close Range Photogrammetry: Principles,
Techniques and Application, p. 528. Whittles Publishing, Dunbeath (2007)
4. Nguyen, H.M., Wünsche, B., Delmas, P., Lutteroth, C.: 3D models from the black box:
investigating the current state of image-based modeling. In: Proceedings of the 20th
International Conference on Computer Graphics, Visualization and Computer Vision (WSCG
2012), Pilsen, Czech Republic, 25–28, June, pp. 115–123 (2012)
5. Baglioni, L., Migliari, R., Salvatore, M.: Calcolo della risoluzione delle riprese panoramiche
delle quadrature piane. In: Valenti, G.M. (ed.) Prospettive Architettoniche: conservazione
digitale, divulgazione e studio, pp. 809–816. Sapienza Università Editrice, Roma (2014)
6. Baglioni, L., Fasolo, M.: Riflessioni sull’illuminazione artificiale di superfici con quadrature.
In: Valenti, G.M. (ed.) Prospettive Architettoniche: conservazione digitale, divulgazione e
studio, pp. 679–684. Sapienza Università Editrice, Roma (2014)
7. Chiabrando, F., Lingua, A., Noardo, F., Spanò, A.: 3D modelling of trompe l’oeil decorated
vaults using dense matching techniques. Isprs Ann. Photogrammetry Remote Sens. Spat. Inf.
Sci. II, 97–104 (2014)
The Application of Photogrammetry on Digitization
and Promotion for Monuments and Temples in Taiwan -
Taking Chua Family Ancestral Temple as an Example
Wun-Bin Yang1,2 ✉ , Tsung-Juang Wang1, and Ya-Ning Yen2

( )
1
National Taipei University of Technology, 1, Sec. 3, Zhongxiao E. Rd., 106 Taipei, Taiwan
{t103859001,tjwang}@ntut.edu.tw
2
China University of Technology, 56 Sec. 3 ShingLong Rd., 116 Taipei, Taiwan
{wunbin,alexyen}@cute.edu.tw
Abstract. In this study, the digitization projects of monuments and temples in

Taiwan were conducted using photogrammetric technology along with 3D Laser
Scanning Point Cloud Data to carry out accuracy assessment of photogrammetry
and 3D laser scanning, and in addition, this study creates a standard operating
procedure for the photogrammetry that can be used to map the cultural relic’s
facade of traditional temples in Taiwan. Moreover, a special workshop was estab‐
lished in order to promote the rapid modeling method for the photogrammetric
technology by means of Chinese language version of CIPA 3 × 3 rules that was
officially introduced at CIPA’s 2015 Annual Meeting, so that this study can
conduct the learning and educational programs to achieve the combination of
cultural assets and public participation.
Keywords: Photogrammetry · Cultural heritage · Temple · Digital techniques ·

Learning
1 Introduction
Normally, the digitization of cultural property is more costly and takes much more time
in terms of instruments and manpower. While comparing LiDAR with photogrammetry,
the cost of LiDAR instrument is the most expensive, it is even up to ten times more
expensive than photogrammetry. The process of post-production of digital images is
more complicated and much more costly as well. On the other hand, these costs usually
are deemed to be the critical factors affecting both the accuracy and usage of digital
images. Thus, how to choose a suitable digital technology to satisfy different require‐
ments has become one of the most important issues in this field.
Theoretically, the staff associated with cultural property wishes to get the most real‐
istic 3d model with greater texture detail, shadowing and decoration effects for the
historic monuments and heritage buildings, for instance, the application of ll3D laser
scanning and digitization. However, the preservation of cultural property is always
limited by the space, time, budget and manpower. It is, therefore, necessary to choose
the suitable technology based on the work characteristics in order to build digital 3D

DOI: 10.1007/978-3-319-48496-9_31
388 W.-B. Yang et al.
models with the best effects within limited resources. Comparing to highly accurate 3D
laser scanning technology, the photogrammetry modeling technology can provide a
rapid modeling method with low cost and low threshold, which is very suitable for
preserving the digital records of historical sites, and hence, photogrammetry modeling
technology has already become the main trend of rapid modeling technology.
There is some degree of difficulty in establishing SOP for various 3D image digiti‐
zation, analyzing or studying various technologies related to the applications of cultural
properties. Additionally, the digitization of cultural property usually takes longer and
costs more than expected. This study, therefore, aims to contribute to the development
of low-cost photogrammetric technology, and meanwhile, solving the problems of low
accuracy and low usage of digital images affected by low-cost equipment. Under this
circumstance, how to choose a suitable digital technology to meet the different requests
from the users has become a bigger issue that needs to be resolved.
Relevant studies for applications of photogrammetry to historical monuments in
Taiwan is still in the developing stage, in which, the other type of tangible cultural
heritage is only a few cases. Main reason is that laser layer technology is popularly used
for developing 3D images by the domestic industry, they seldom use the mature photo‐
grammetric technology introduced by the foreign industry. However, in recent years,
due to the high expenditure of advanced 3D laser scanning equipment and difficulty in
operation due to its large dimension in addition to higher technology threshold, and this
situation makes it impossible to be developed in Taiwan. Thus, this study would like to
introduce photogrammetric technology to establish low-cost digital imaging technology
to be used particularly for historic monuments.
Over the past several decades, Taiwan’s rapid economic development totally ignored
the importance of cultural preservation. At present, the national consciousness has urged
the people in the whole country to take on the responsibility for preserving cultural
properties. To enhance the local perspectives on conservation of cultural heritage, except
the promotion programs sponsored by the government, actual activities involved in
dynamic preservation should be conducted as well, which should be in conjunction with
school education to create the community sense and public participation; it is hoping
that we can achieve the sustainability and management of cultural properties through
new digitization technology. To enhance and popularize the perspectives of digitization
and preservation of cultural heritages, this study promotes the low-cost digital imaging
technology used for recording historic monuments through the cultural activities held
by a special workshop, aiming to accomplish the basic task for the promotion of educa‐
tion for cultural properties digitization.
2 Background
Photogrammetric technology contains aerial photogrammetry and close-range photo‐

grammetry. The coverage area of the image developed by the aerial photogrammetry
normally is large, which is suitable for reconstructing images for a large number of
buildings in the large area. However, the photographing was taken vertically and moat
of images were overlapped, the covered condition was more severe and it was hard to
establish detailed models. In general, the distance for photogrammetry is limited to less
The Application of Photogrammetry on Digitization and Promotion 389
than 300 m. The camera station can be set up on the ground or on a high tower. Having
a good planning in advance, it still can take images for the building facades from every
aspect and it will also completely cover each detailed structure of the buildings. So that
it can build a detailed model for the buildings. Close-up images even can be used as the
real image materials for posting the surface of buildings in order to build more realistic
virtual reality scenes for buildings.
The earlier photogrammetry application used for historic monuments in Taiwan was
the measurement for engineering drawings (Chien 1984). As to the study of 3D digiti‐
zation of ancient temples and historic monuments, it is still in the phase of initial devel‐
opment. At present, relevant studies and applications related to the photogrammetric
technology are growing towards maturity (Ioannides et al. 2013; Doulamis et al. 2015).
The applications of 3D digitization for historic monument objects can be summarized
as follows:
Currently, the automatic technology of photogrammetry is very popular, and 3D
technology used for mapping the buildings having overall uncomplicated, geometric
appearance along with fewer details is far more mature than before, which has signifi‐
cantly reduced the technology threshold and costs in the digitization and physical pres‐
ervation of cultural properties. With a good method, so that ICT technology can be used
to bridge the gap between cultural heritage and learning (Ott and Pozzi 2011). Further‐
more, Lower threshold for photogrammetric technology in recent years has increased
the level of affinity that will attract more users. While combining the power of public
participation, it will effectively enhance the possibility of the digitization for cultural
properties. In the future, photogrammetric technology will become a practical method
for on-site digital recording of activities when the public is visiting the cultural proper‐
ties. (De Reu et al. 2013; McCarthy 2014) (Fig. 1).
Fig. 1. Elevation and photographic 3D model of the minaret of the Qala of Beni Hammad
(Almagro 2013; Almagro et al. 2015) and The documentation of traditional Batak Toba House
(Hanan et al. 2015)
3 Digitization Work and Procedures
The subject of this study was Chua Family Ancestral Temple at the Chiung Lin Village,
Kinmen County, Taiwan, which was built in 1840. It is an architecture of courtyard
house with two-row and two-river and it is a national historic monument specified by
the Taiwan government. The building materials were very elegant. In which, the roof
of the main hall surpasses all other family temples in height in the same village and it
has become one of the characters of Chua Family Ancestral Temple. At present, this
temple is the only family temple in Chiung Lin area, which still keeps its construction
monument. The building plane was adopted traditional architectural pattern, combining
with obvious main axis and pattern in bilateral symmetry. The overall layout of the
temple is developing and creating many flexible, smooth & internal spaces between
reality and falsehood.
3.1 3D Laser Scanning

In this study, the 3D laser scanning technology was used firstly to establish overall
coordinate system around the object. Then, the point cloud data generated by 3D laser
scanning were used, based on original coordinate system and through specified loca‐
tions, to relocate to a pre-positioned coordinate system in order to improve the accuracy
of measurement. Moreover, after conducting the calibration of 3D point cloud coordi‐
nates, a 3D point cloud model was created, which could be used as the basis data and
value-added application for performing the follow-up comparison and analysis.
a. 3D Outdoor Laser Scanning of Chua family Ancestral temple
This study used Riegl LMS‐Z420i scanner to implement digitized record for the
outdoor of Chua Family Ancestral Temple. The scanning area contained the whole
area on all sides. The marked points of long-range scanner could be classified into
adhesive type and cylindrical type. In this study, ten ground scanning stations had
been established in the area outside the Chua Family Ancestral Temple, and wherein,
two stations were set up on each side (left and right) of building elevation at the
highest point of roof; and moreover, there were 17 marked points deployed sepa‐
rately at the various positions either on the ground or the wall so as to build the
marked point system that could improve the efficiency of integrating accuracy and
point cloud data (Fig. 2).
b. 3D Indoor Laser Scanning of Chua Family Ancestral Temple
Its interior space was scanned with a FARO Laser Scanner Focus3D. The marked
points for the short-range scanner could be classified into the adhesive type and
cylindrical type. This study established fifteen indoor scanning stations along with
ten marked points that were deployed separately at the various positions either on
the ground or the wall in order to build the marked point system (Fig. 3).
c. Results of 3D Scanning and Digitization of Chua Family Ancestral Temple
In this study, the specific software was used for integrating the results of indoor point
clouds with outdoor point clouds, which could be used as the basic information for
implementing the digitization analysis of this study. The results are as follows (Fig. 4).
Fig. 2. Outdoor scanning work
Fig. 3. Indoor scanning work
Fig. 4. Results of indoor and outdoor point cloud, and CAD drawing
3.2 Digitized Photogrammetry

The operation of photogrammetry was implemented according to the Chinese-language
version of CIPA 3 × 3 rules officially introduced at CIPA’s 2015 Annual Meeting.
Wherein, the cameras were set up vertically by facing the wall, and the cameras with
featured members, which could be used for the photos overlapping, were moved from left
to right along a horizontal track. At the corner of the wall, multiple photos should be taken
additionally in order to set up the connection to transfer information between two walls.
As for the ground, because the ground of Chua Family Ancestral Temple is paved
with bricks matching of repeated patterns, the difference between bricks is very little.
While considering a higher probability of success at the stage of follow-up data
processing, hence, the wide-angle lens were used to shoot the photos, because one single
photo can cover a much larger area. The porch of the Chua Family Ancestral Temple is
very narrow and it seems difficult to take better photographs via standard operation
procedure. Under this circumstance, incomplete data may arise with regard to the inner
portion of porch floor grating.
In the case study of Chua Family Ancestral Temple conducted in this year. While
computing with about 1200 photos on the same computer, total running time for using
VisualSFM is about 250 h, instead, Photoscan runs about 100 h only. Photoscan has a
built-in module that can be used for the Mesh modeling. However, VisualSFM should be
integrated with CMPMVS prototype. After testing, it was found that modeling function of
CMPMVS could be applied to small objects only. It was not suitable for the building
structure contained in the monuments. However, the mesh modeling function of Photo‐
scan had a better integrity and applicability for such kinds of buildings (Fig. 5).
Fig. 5. Digitization of Wind-lion Gods at Chua Family Ancestral Temple
4 Comparison and Analysis for Digitization Results
This study used the images of scenes shot at the Chua Family Ancestral Temple and
inputted them into Photoscan for scene reconstruction in order to build point cloud
models. In addition, point clouds measured by 3D laser scanner were used as a basis for
assessment for evaluating the point clouds. While comparing the point clouds of Photo‐
scan with the point clouds of LiDAR, the coordinate system defined by each tool was
belonging to a relative coordinate system, moreover, there is no any control point added
in the procedure of point cloud establishment in Photoscan to be a scale basis, therefore
the first task before starting to evaluate the point clouds was to transfer the point clouds
of two groups into the same coordinate system. Thus, this study adopted the coordinate
system and scale of LiDAR as a normative reference. So that the point cloud models
generated by Photoscan were combined with the point could models generated by
LiDAR and then all point cloud models would be transferred into the same coordinate
system defined by LiDAR.
As to the accuracy analysis of point clouds, this study used the way of a fitting plane
to evaluate the accuracy of 3D point clouds generated by two groups. On the flat surface
such as wall or ground, point clouds of both groups would be cut synchronously so as
to extract the point clouds of such flat surface. Also, the point clouds of LiDAR would
be used as the fitting point of the said plane to find out the plane equation of such flat
surface. After obtaining the plane equation, the deviation between the Photoscan’s point
clouds and such flat surface would be calculated in order to evaluate the accuracy.
In the part of the architectural appearance, it is divided into the outer wall (left, right
and rear) and the courtyard house area. The extracted plane is shown in relevant figures,
in which, it is divided into 13 plane areas. The total point clouds of each area are listed
in Table 1. According to Table 1, it can be found that the number of point clouds gener‐
ated by Photoscan is greater than the number of point clouds scanned by LiDAR. As a
result, the texture of point cloud models can be seen clearly (Fig. 6).
Table 1. Number of point clouds

Partition of point cloud Number of point cloud
LiDAR Photoscan
S1_Left 98,982 4,402,590
S2_Right 216,568 3,104,981
S3_Back 125,423 1,963,125
1–1 3,784 133,034
1–2 3,802 107,779
2 1,004 275,643
3 1,662 459,100
4 14,804 1,430,267
5–1 703 29,492
5–2 998 29,251
6 912 68,865
7–1 795 94,283
7–2 677 96,329
Table 2 contains the accuracy analysis and the plane parameters of the fitting plane for
each area obtained by means of LiDAR’s point clouds. It can be found in Table 2 that the
accuracy of the fitting plane of LiDAR is about 1 cm. However, on the partition of point
cloud 1–2, a bigger flat error will appear, about 3.4 cm. While using the Photoscan’s point
clouds in such area to calculate the error of point clouds against the fitting plane, the most
Fig. 6. Accuracy analysis of point clouds
errors are less than 5 cm. On the rear wall, there is a bigger error that is about 11 cm. There
might be two reasons for such errors. One reason is that the error is produced by the point
cloud, the other reason is that the error is occurring during the transferring procedure of the
coordinates. In general, the average accuracy of the fitting plane of point clouds of Photo‐
scan in the external area is about 3 cm.
Table 2. Accuracy analysis of point clouds (RMSE Unit: M)

Exterior LiDAR (Surface equation Ax + By + Cz + 1 = 0) Photoscan
Partition of point cloud A B C RMSE RMSE
S1_Left −0.1768 −0.2011 0.0074 0.0136 0.0407
S2_Right −0.0407 −0.0463 0.0022 0.0100 0.0249
S3_Back 0.0236 −0.0207 0.0001 0.0118 0.1168
1–1 0.0832 −0.0732 −0.0039 0.0104 0.0028
1–2 0.0835 −0.0731 −0.0045 0.0344 0.0139
2 0.0559 −0.0333 0.3835 0.0199 0.0323
3 0.0480 −0.0171 0.5829 0.0160 0.0262
4 0.0216 0.0061 0.5845 0.0092 0.0329
5–1 0.0786 −0.0720 −0.0019 0.0081 0.0022
5–2 0.0825 −0.0730 −0.0019 0.0077 0.0172
6 −0.0416 −0.0474 0.0021 0.0072 0.0591
7–1 0.0622 −0.0555 −0.0021 0.0069 0.0118
7–2 0.0631 −0.0553 −0.0005 0.0082 0.0067
AVG 0.0126 0.0298
5 Comparison Analysis of Costs of Producing Photogrammetric

Products
The discussion related to photogrammetry cost will be divided into three portions:
recording time of photos, data processing time (office work) and computer calculation
processing time. Table 3 lists the real-time statistics information for shooting photos of
the Chua Family Ancestral Temple during the photogrammetry operation.
The photos were shot by photogrammetric professionals who spent a whole day to
complete his work. The light and climatic conditions on the site must be considered.
The photo shooting on the site implemented according to SOP can reduce the time of
field photography and data processing. However, existing SOP used for ancient monu‐
ments as well as middle and large scale buildings still needs to be improved. Thus, it is
necessary to improve the SOP for photographing.
Table 3. Photogrammetry mapping work (Unit: hr)

Project Photo shoot (Field) Office work Computer
Professionals General Professionals General processing
Experimental 16 0 100 0 500
SOP work 4 4 24 0 200
6 Education and Promotion
CIPA 3 × 3 rules were established by Peter Whaldausl and Cliff Ogleby in 1994. The
rules used photographic images as a feasible method for the implementation of meas‐
urements and applications in the future. The newest version (edition Nov. 2013) was
updated by Bill Blake, Andreas Georgopoulos and José Luis Lerma (News posted in
CIPA website). CIPA has compiled and produced “The Photogrammetry Documents
related to the Simple Structure of 3 × 3 Rules” acting as a simple digitization method
for cultural monuments and historical sites. The development principle of CIPA 3 × 3
can be applied to an easy and convenient photogrammetry method, which should be
vigorously promoted in the domestic market.
The newest Chinese language version of CIPA 3 × 3 rules with additional correlation
contents was officially introduced at CIPA’s 2015 Annual Meeting. So that this study
will use such rules to teach students and carry out simple training program of photo‐
grammetry along with on-site operations in order to achieve the outcome of the promo‐
tion of low-cost digitization of cultural properties (Fig. 7).
Fig. 7. Photogrammetric workshop and on-site operation
7 Conclusions and Future Suggestions
This study aims to explore the feasibility, rough cost estimation, and initial accuracy
analysis of photogrammetry that can be used for low-cost digital preservation of the
monuments, historical buildings and ancient temples in Taiwan. After testing, the
existing close-range photogrammetric operation procedures has good effects for small-
middle sized objects or geometry objects, such as stone lion, stone carving and small
pavilion. However, it is still necessary to further design SOP for middle-large sized
buildings with curving shaping. Generally speaking, the higher accuracy of photogram‐
metric products, the higher operating costs. Thus, it needs to choose the most suitable
equipment with appropriate accuracy level to meet the requirements as well as economic
costs and benefits from various users. Based on the perspectives of preservation of
monuments in Taiwan, the subsequent studies in the future, including: 1. Explore the
simple and high precision photogrammetry techniques used for different types of monu‐
ments in Taiwan, and also compare the differences between each other; 2. Analyze the
influence of adding control points, lights and shadows to same materials on the accuracy
and integrity of different types of monuments; and 3. Study on the adaptability of photo‐
grammetry combining with the public participation.
In the future, the possibility of using Mesh Model in the measurements and appli‐
cations of historic monuments repair can be further discussed. The maturation and auto‐
mation degree of photogrammetry will enhance the possibility of crossing the high
threshold of digitization technology while carrying out digitization works and enhance
the opportunity of public participation.
Acknowledgements. This study is sponsored by the Bureau of Cultural Heritage, Ministry of

Cultural for 2015, 2016 Research and Development Project Plan (104-4, 105-3).
References
Almagro, A.: Surveying world heritage Islamic monuments in North Africa: experiences with
simple photogrammetryic tools and no previous planning. In: ISPRS Annals of the
Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. II-5/W1 (2013)
Almagro, A., Tandon, A., Eppich, R.: First aid to cultural heritage: training initiatives on rapid
documentation. In: The International Archives of the Photogrammetry, Remote Sensing and
Spatial Information Sciences, vol. XL-5/W7 (2015)
Chien, H.K.: The Application of Historic Preservation by Photogrammetry, Master thesis. NCKU
(1984)
De Reu, J., Plets, G., Verhoeven, G., De Smedt, P., Bats, M., Cherretté, B., DeMaeyer, W.,
Deconynck, J., Herremans, D., Laloo, P., Van Meirvenne, M., De Clercq, W.: Towards a three-
dimensional cost-effective registration of the archaeological heritage. J. Archaeol. Sci. 40(2),
1108–1121 (2013)
Doulamis, A., Soile, S., Doulamis, N., Chrisouli, C., Grammalidis, N., Dimitropoulos, K., Ioannidis,
C.: Selective 4D modelling framework for spatial-temporal land information management system.
In: Third International Conference on Remote Sensing and Geoinformation of the Environment,
p. 953506. International Society for Optics and Photonics, June 2015
Hanan, H., Suwardhi, D., Nurhasanah, T., Bukit, E.S.: Batak Toba cultural heritage and close-
range photogrammetry. Procedia Soc. Behav. Sci. 184, 187–195 (2015)
Ioannides, M., Hadjiprocopi, A., Doulamis, N., Doulamis, A., Protopapadakis, E., Makantasis,
K., Julien, M.: Online 4D reconstruction using multi-images available under Open Access.
ISPRS Ann Photogramm. Remote Sens. Spat. Inf. Sci. 5, W1 (2013)
McCarthy, J.: Multi-image photogrammetry as a practical tool for cultural heritage survey and
community engagement. J. Archaeol. Sci. 43, 175–185 (2014)
Ott, M., Pozzi, F.: Towards a new era for Cultural Heritage Education: discussing the role of ICT.
Comput. Hum. Behav. 27(4), 1365–1371 (2011)
3D Acquisition, Processing and Visualization of
Archaeological Artifacts
The Samarra Collection of the Museum
of Islamic Art in Berlin
Arie Kai-Browne1 ✉ , Kay Kohlmeyer1, Julia Gonnella2, Thomas Bremer1,

( )
Susanne Brandhorst1, Felix Balda1, Sebastian Plesch1, and Dennis Lehmann1

1
Landscape Archaeology, Game Design, Computer and Media Sciences, University of Applied
Sciences Berlin, Berlin, Germany
{Arie.Kai-Browne,Kay.Kohlmeyer,Thomas.Bremer,
Susanne.Brandhorst,Felix.Balda,Sebastian.Plesch,
Dennis.Lehmann}@HTW-Berlin.de
2
Museum für Islamische Kunst Berlin (SMB), Berlin, Germany
[email protected]
Abstract. In the past decade there has been a steady increase in research projects
dealing with the three-dimensional documentation of cultural heritage. While 3D-
scanners and photogrammetry are widely used for documenting historical monu‐
ments and archaeological excavations, the application of this technology within
museums has not yet been established within the daily work routine. Even though
the benefits of 3D-documentation are quite manifold, usually only outstanding
artifacts are being recorded in this manner due to the complex workflows for
deriving datasets, which can be used for further research and knowledge transfer.
The interdisciplinary research project MOSYS-3D has been dealing with the
entire workflow ranging from data acquisition, pre- and postprocessing steps as
well as testing different forms of visualizations.
1 Introduction
In general, museums and other institutions dealing with archaeological artifacts have
been reluctant to integrate a 3D-digitization pipeline into their daily workflow due to
the fairly complex and time consuming task of acquiring, processing and visualizing
high resolution 3D-data. Even though the benefits of 3D-documentation are manifold
for scientific research as well as educational purposes, usually only special finds are
being digitally recorded in three-dimensions while large collections are mostly recorded
using classical two-dimensional approaches.
The range of possibilities for the further use of high resolution 3D-data includes a
variety of applications in the field of restoration/conservation, such as deformation
analysis [1–3] the construction of fillings or supporting elements [4], or as the basis for
condition mapping and archiving restoration information [5]. Of course, it is possible

DOI: 10.1007/978-3-319-48496-9_32
398 A. Kai-Browne et al.
to use rapid prototyping for creating replicas [6] or positives of molds [6]. For further
scientific research 3D-data can be visualized in a standardized way, enabling the compa‐
rative study of shapes [7], while it is also possible to visually enhance fine structures,
such as tool marks [8], fine carvings [9] or inscriptions. Furthermore the 3D-data can be
manipulated, enabling the unwrapping of cylindrical objects, such as cylinder or similar
artifacts [10]. Other interesting applications for scientific research include the automatic
detection of cuneiform scripts [11], where the methodology itself is transferable to
similar research question. Also automatically fitting large numbers of digitized single
fragments [12, 13] offers a major benefit for the practical application of restoring frag‐
mented artifacts.
The interdisciplinary research project “MOSYS-3D” (Mobile, modular System for
highly accurate 3D Documentation of Cultural Heritage) is a cooperation of the Univer‐
sity of Applied Sciences Berlin (HTW) and The Museum of Islamic Art Berlin (SMB),
bringing together archaeologists, electronic engineers, game designers and computer
scientists. The research aims at tackling different aspects of 3D-documentation, ranging
from the efficient acquisition of 3D-data, automating pre- and post-processing steps and
finding new ways for visualizing these data sets for knowledge transfer and scientific
research questions.
We believe that utilizing interactive 3D-environments, such as game engines, offers
manifold possibilities for the further usage of acquired 3D-data. The immediate inter‐
action between the user and the digital data offers a new dimension of comprehending
complex three-dimensional situations. Such environments do not only serve the purpose
of displaying single 3D-models, but also enable the combination of different datasets,
which have been derived from a multitude of different platforms with varying spatial
scales, different resolutions and different accuracies. This allows, for example, visual‐
izing small-scaled artifacts within their immediate archaeological context, such as the
excavation area. Furthermore, the excavation site itself can be embedded within its
surrounding landscape. Hence a combination of spatial scales can be achieved, which
range from sub-millimeter to kilometers [14].
But for combining large amounts of 3D data, which have been derived from 3D-
scanners, it is necessary to drastically reduce the file size, while retaining a certain degree
of detail. Following we will show the necessary steps for efficiently processing high
resolution data for combining large amounts of objects within one common environ‐
ment, such as game engines or for web-based presentations.
2 Case Study - The Ancient City of Samarra
Samarra, located about 125 km north of Baghdad, is one of the outstanding sites in
Islamic art history and archaeology. Here the first, more or less, systematic excavations
took place in the years 1911–1913 under the supervision of Ernst Herzfeld. The Abbasid
city of Samarra, extending over an area of 57 km2, is one of the largest archaeological
sites of the ancient world. Between 836 AD and 883 AD it was the administrative center
of the Abbasid caliphate, from which the most significant and largest empire in Islamic
history was governed. The Islamic culture and scientific achievements of this period had
3D Acquisition, Processing and Visualization of Archaeological Artifacts 399
a lasting effect on later developments in Europe. Large parts of the Samarra finds exca‐
vated by Ernst Herzfeld were brought to the Kaiser-Friedrich-Museum Berlin and are
currently in the Museum of Islamic Art Berlin. The finds include over 90 ornamented
stucco panels, painted wood fragments, outstanding glass ware, pottery and stone arti‐
facts. Unfortunately the current exhibition does not enable the observer to truly under‐
stand the importance and the historical context of the finds, therefore counteracting the
actual purpose of a museum: Enhancing the understanding for our cultural heritage.
Hence the Samarra finds in the Museum of Islamic Art in Berlin offer an ideal test case
for demonstrating the potential but also the challenges of high resolution 3D-documen‐
tation, the further scientific analysis and possible ways of visualizing these datasets for
knowledge transfer.
Therefore one aspect of the MOSYS-3D project was to digitize a large part of the
Samarra collection at the Museum for Islamic Art Berlin, as groundwork for the virtual
contextualization of the artifacts. For this purpose a vast array of historic photographs,
excavation plans and Herzfeld’s field notes have been collected and analyzed, enabling
the exact localization of the digitized stucco panels within single rooms of different
buildings. For the contextualization of the digitized finds, the archaeological excavation
plans of Ernst Herzfeld form the groundwork for the complete three-dimensional recon‐
struction of selected buildings. For combining the reconstructed buildings and the stucco
panels a game-engine serves as a common environment, where the final visualization
can be realized in a way, that it enables easy access to researchers as well as laymen.
3 Methodology
Within the field of 3D-digitization of cultural heritage the applied methodology can vary
immensely depending on the research questions at hand, the spatial scales involved, the
type of objects to be recorded, the technical equipment used and so on [15, 16]. Further‐
more the different types of 3D-digitizing technology applied require different
approaches in regards of data acquisition and processing, therefore making it necessary
to develop appropriate workflows to achieve the desired results.
Generally, data acquisition can be classified into active and passive techniques.
Active methods, like 3D-scanners, rely on a self-induced signal to acquire spatial data
while passive methods, such as photogrammetry or multi-spectral imaging, use ambient
light. Both types of data acquisition strategies have their advantages and disadvantages,
making it necessary to precisely define the cases in which one method should be
preferred over another or if maybe the combination of both methods is beneficial [17].
In the case of digitizing the artifacts at the Museum for Islamic Art Berlin, it was deter‐
mined that an active method of data acquisition would be necessary to ensure a ubiqui‐
tous data quality in regard to the accuracy. For a large extent the surfaces of the artifacts
have a diffuse reflection, which is ideal for 3D-scanning as well as photogrammetry. But
some areas do not have a sufficiently varying textured or structured surface, which can
lead to an increase in noise when applying Structure-from-Motion based algorithms.
Also, dealing with 3D-aqcuisition methods, it is crucial to evaluate the functional

principle, which determines whether the resolution, accuracy and range are sufficient
for the research question at hand [18–20].
3.1 Data Acquisition

Hardware. For digitizing the Samarra finds the close-range 3D-scanner breuckmann
smart SCAN3D-HE C5 was utilized, which is based on the fringe projection technique
and uses two asymmetrically arranged 5 MP CCD-cameras for the data acquisition [21].
For each pixel the depth information is calculated, resulting in five million 3D-meas‐
urement points per scan. For the final computation of the 3D-data, it is possible to sepa‐
rate the input data according to the used camera/light-projection combination. The
highest accuracy is achieved by using the data visible to both stereo-cameras (30° trian‐
gulation angle), though less surface of the object can be captured compared to using
only a single camera and the light-projection. The distance between the object-of-interest
and the 3D-scanner is static due to the functional principle, so only three different
working ranges are available for this specific system. Given the static range between the
object-of-interest and the scanner, the final spatial resolution of the 3D-scans is deter‐
mined by the field-of-view of the lens. Therefore the lenses S-125, M-475 and L-950
were acquired, as a trade-off to capture small and large objects with a sufficient resolution
in an appropriate time-frame. For most artifacts, including the stucco decoration, the
3D-scanner was used in combination with the M475 measurement field, delivering a
spatial resolution of 0.17 mm and an accuracy of ±0.035 mm [21] (Fig. 1).
Fig. 1. Typical result of 3D-scanning; very accurate, high resolution mesh
Data Acquisition Strategies. The data acquisition strategies for digitizing artifacts
using close-range 3D-scanners can vary to a certain degree. In general, each single 3D-
scan represents one viewpoint, therefore making multiple, overlapping scans necessary
to capture the entire surface of an object [22–24]. To acquire all necessary viewpoints,
either the 3D-scanner itself or the artifact has to be re-positioned for every consecutive
scan. Whether the 3D-scanner or the artifact has to be re-positioned depends mainly on
the dimension and the fragility of the artifact. The individual viewpoints need to be
aligned to create a single 3D-model of the object. There are diverse options on how to
align all the single 3D-scans within the same coordinate reference system, which influ‐
ences the data acquisition strategy. Nonetheless a certain amount of overlap between
two scans is required for computing a best-fit match based on the geometrical features
of the object.
One of the most common ways to align the individual scans is to manually define at
least three overlapping correspondence points of two range maps, which act as a starting
reference for a finer alignment based upon best-fit algorithms such as the well-known
ICP (Iterative Closest Point) algorithm [25] (Fig. 2). This alignment step is usually part
of the data acquisition process and is conducted after each single scan, adding a fairly
significant increase of time required for capturing the entire object.
Fig. 2. Typical alignment of single scans using an ICP-based approach
Another possibility is the use of dedicated markers, which have to be distributed on

or around the object-of-interest. These markers are automatically recognized by the
scanning software, negating the need for manual definition of correspondence points
[26]. Using this approach, it is necessary to keep the markers and the object-of-interest
in the same orientation relative to each other. This method will not work with larger
scaled, unmovable objects, such as the stucco decoration, where sticking the markers
onto the objects surface cannot be done due to conservational considerations. Further‐
more systems have been developed, which use an optical tracking system [27, 28] or
that try to automatically align the 3D-scans based on texture features and geometrical
properties of the object [26].
The approach of re-positioning the 3D-scanner on a tripod, especially when dealing
with a large number of artifacts, becomes quite time consuming. Hence another part of
the MOSYS-3D project was to develop a system for the efficient acquisition of 3D-data
of unmovable objects. For this purpose a combination between a robotic system and the
3D-scanner was developed. The 3D-scanner was mounted onto the 6-axis robot arm pi4
UR10, which has a working range of 1300 mm and can record the position and the
orientation with a repeatability of ±0.1 mm [29]. The positional information of each
viewpoint is transferred to OPTOCAT 2015, where a macro was used to place the single
scan within a common coordinate system, negating the need for additional markers or
manual placement of correspondence points.
Color Information. In addition to the spatial data, the 3D-scanner is capable of

capturing color information. The color information is recorded with the same imaging
sensor as the depth information; hence an 8-bit RGB value is assigned to every 3D-point
measurement. Due to the fairly small sensor size, the captured color information can
show a fairly large amount of image-artifacts, which can be minimized through the
possibility of capturing and averaging multiple color images [26].
For achieving high quality color data the lighting setup is essential. Preferably soft
lighting should be setup in a way to minimize hard shadow casts and allow a uniform
illumination of the object. This can easily be achieved when digitizing small finds under
lab conditions but in case of the MOSYS-3D project the stucco panels, which are
mounted to the walls of the depot and in the museum itself, the lighting situation was
not controllable due to the changing natural lighting coming through large windows. It
was not feasible to move the stucco panels to a controlled lab situation because the
efforts to do so would have been out of the project’s scope. But a major problem arises
when using the ambient lighting for capturing color information: the shadows are baked
into the colors, making the virtual relighting troublesome. Since the shadows are already
captured within the color information, one is not able to virtually relight the scene
because the direction of the virtual lighting and the captured lighting can be contra‐
dicting.
Therefore, additional photographic images were captured using the full-frame DSLR
Nikon D600 in combination with the Sigma Art DG/HS 50 mm 1,4f lens. A ring flash
was used in combination with appropriate exposure times to suppress the ambient
lighting and to minimize shadows in the images. This approach enabled a uniform data-
set of all stucco panels, enabling the comparative study of the color information.
3.2 Processing
Post-processing. The post-processing steps are crucial for deriving a continuous, high-
resolution mesh from the acquired 3D-scans, which serves as basis for further analysis
and visualizations. After the data acquisition the 3D-scans are already triangulated,
negating the step of meshing a point-cloud. During the acquisition one already has to
define certain parameters for the final computation of the depth-maps, for example
whether only the data from the stereo-cameras should be used or whether data from all
possible camera-light projector combinations should be used. This can be recalculated
again in the processing phase, which can be especially useful when difficult surfaces,
such as shiny materials, have been recorded. The individual viewpoints of the 3D-scans
usually are already roughly aligned, following the previously mentioned acquisition-
strategies. Nonetheless a global registration of all range maps is necessary to reduce the
overall alignment error.
Quite often a manual clean-up of unwanted or unnecessary data, such as stray points
or objects in the background is needed, which - depending on the content and acquisition
strategy - can be automated to a certain degree. OPTOCAT offers the possibility to
automatically mask dark areas with a definable threshold, so if one uses a dark surface
for placing the artifacts, this area will be automatically excluded. It is also possible to
take a scan of the background area and use this as a mask. After cleaning the unwanted
parts of the 3D-data the single scans can be merged into a single, continuous mesh.
Within OPTOCAT it is possible to define whether the mesh is supposed to be “water-
tight”, meaning that all scan holes with a defined size will automatically be filled.
Furthermore it is possible to decimate the mesh according to different filter parameters,
such as the maximum spatial deviation to the original scan. After the final mesh has been
computed, quite often additional repair work is necessary to eliminate self-intersecting
triangles, small outliers, tunnels and spikes. This task was done using the Geomagic
Studio Pro 12 software, where the mesh consistency is automatically analyzed and the
mentioned errors repaired. After this step, the high resolution mesh can be exported to
typical 3D-data formats, such as .ply or .obj, which can easily be imported to other 3D-
modeling or visualization software.
For texturing the 3D-scans with external, high-quality imagery it is necessary to first
do a photogrammetric reconstruction of the acquired digital images. The reconstruction
is based on a common Structure-from-Motion/Multiview-Stereo approach implemented
in the Agisoft Photoscan 1.2 software. The 3D-scan is subsequently aligned to the
reconstructed dense point cloud using an ICP-approach implemented in the open-source
software Cloudcompare [30]. The 3D-scan can then be imported to Agisoft Photoscan,
where it is possible to use the computed camera positions to project the color information
of digital imagery onto the mesh of the 3D-scan (Fig. 3).
Fig. 3. Combination of 3D-scanning and photogrammetry
Further Processing. Following the typical post-processing steps for deriving a high
resolution, textured 3D-modell of the artifacts, various methods used in the field of
gaming technology have been employed for drastically reducing the file size while
maintaining a very high degree of detail. This is a necessary step for integrating large
amounts of 3D-objects into an interactive real-time environment or for enabling the
exchange of 3D-data via the web.
In the first step the high resolution mesh is decimated to approximately 1 % or less
of its original spatial resolution. The decimation algorithm in OPTOCAT takes the
curvature of the mesh into account, so flat surfaces will be decimated more than curved
areas. After the decimation step, fine surface details, such as tool marks or fine carvings,
are lost, and only the basic shape of the object is retained. For retaining the fine details,
different texture maps have to be created, such as normal-, diffuse-, ambient occlusion-
and curvature maps. Mapping textures i.e. pixel information, onto a mesh requires the
computation of a UV-map, which means the mesh consisting of polygons has to be
unfolded onto a two-dimensional plane. The UV-coordinates ensure the spatial link
between the x, y, z coordinates of the mesh polygons and the image. After the UV-
mapping the different textures can be computed, which is based on projecting surface
information from the high resolution mesh onto the triangles of the decimated mesh,
converting geometrical characteristics into pixel information. For example, for fine
details a normal map contains the surface normal vector of the high-resolution mesh,
which is stored as a RGB value in the pixel image. The RGB value represents the x, y, z
value of the surface normal vector. This enables the interactive lighting of fine surface
details. Ambient-occlusion maps store self-shadowing information, while curvature-
maps define concave and convex areas of the mesh. When combined, these maps enable
a realistic appearance of the decimated mesh, which visually is hardly distinguishable
from the high-resolution mesh.
A major disadvantage of these post-processing steps is that for achieving optimal
results a multitude of different software is required to be used. For example, the deci‐
mation can be done either in the manufactures software or, sometimes even with better
results, in other software such as Meshlab, Geomagic Studio and many more. For UV-
mapping a range of dedicated software, such as Ultimate Unwrap 3D, UVMapper or
Headus UV Layout, are available. The computation of the texture maps again requires
other software such as Xnormals or Blender. Therefore this workflow is quite labor-
intensive, requiring a lot of manual intervention.
To reduce the time-frame required for drastically reducing high-resolution 3D-
models while maintaining the fine details, a script was developed for Autodesk 3D
Studio Max 2014 using MAXScript, which automates all the above mentioned steps.
The script can be accessed via a user interface, where multiple 3D-models can be selected
and the type of desired output can be defined. The selected data-sets are imported into
3D-Studio Max, where a modifier, which is a reversible modification of the mesh, is
applied for decimating the mesh. After reducing the polygon count of the mesh, the
model is automatically aligned to the coordinate axis, where an in-house developed
algorithm tries to determine the correct orientation of the object. Afterwards, the
MAXScript activates a Microsoft Windows batch script, which launches Z-Brush soft‐
ware and multiple scripts (Zscripts) for automatically creating a UV-Map. Following
the unwrapping phase, the 3D-model is exported back to 3D-Studio Max, where another
script is launched, which exports the 3D-model to Xnormals. In Xnormals the high- and
low-resolution meshes are imported and the different maps are computed automatically
based on the presets defined in the UI “Scan-Data Archive Helper”. At the end of these
processing steps, the finished decimated model is exported to a typical 3D-data format
and can be used for a variety of purposes, ranging from web-based applications to the
visualization within a game-engine.
This fully automated reduction process enabled to save an incredible amount of time.
Manually processing a single, high resolution mesh took at least a couple of hours,
depending on the complexity of the scanned object. Through the developed automated
workflow described above, the time necessary for deriving a decimated mesh with
different types of texture maps was reduced to around 10–15 min (Fig. 4).
Fig. 4. Visual difference between the high-resolution mesh and the decimated, textured mesh.
Right: 9.700.000 polygons, filesize (.ply): 178 MB - Left: 30.000 polygons, filesize (.ply):
736 KB + 2 textures (.jpeg): 22 MB
4 Conclusions
Within the project MOSYS-3D different key aspects of 3D-documentation, ranging from
the efficient acquisition of 3D-data, automated post-processing steps, and the final visu‐
alization, have been tackled within the case study of the Samarra collection at the
Museum for Islamic Arts Berlin.
Regarding the acquisition of 3D-data, the use of a robotic arm for automatically
positioning single 3D-scans turned out to be not as efficient as initially planned. This is
due to different factors, ranging from the communication link between the proprietary
software for controlling the scanner and the robotic arm as well as the time needed for
eliminating vibrations caused by the movement of the robotic arm. Furthermore, the
depot, where the artifacts and stuccos are stored, does not offer a lot of room for posi‐
tioning the device and many objects are mounted fairly high on the walls, thus are not
reachable. For a more efficient data capture some developments of the scanner-vendor
are very promising. This includes reducing the time necessary for the actual scanning
process as well as a recently implemented feature in OPTOCAT 2016, where the single
scans are automatically aligned based on feature detection and depth information [26].
Significant achievements could be made in the area of processing the 3D-data. The
time necessary for the data reduction and preparation for the use within game-engines
or web-based viewer was reduced drastically, while still retaining high accuracy and
fine details, which can be captured by modern close-range scanners. The manual labor
was reduced to a fraction of the time spent before. The developed workflow enables an
easy way to automatically reduce hundreds of objects with hardly any manual interfer‐
ence.
Furthermore, the benefits of utilizing real-time environments for displaying, inter‐
acting and contextualizing 3D-data has proven to be an efficient way of tackling certain
research questions; some of which first emerged after combining various data sources
within one environment. This also represents an ideal platform for knowledge transfer,
enabling laymen to understand complex hypotheses and appealing especially to the
younger generation, who are quite often rather reluctant to visit museums.
The use of real time environments for comprehending complex, multidimensional
scenarios within a digital environment has tremendous potential and we believe this
approach will be the future of working with 3D-data.
Acknowledgement. We want to express our sincere gratitude for the funding by the European
Regional Development Fund (INNO 02-08/EDV. Nr. 2.1.6_0208_E) and the Regierender
Bürgermeister von Berlin, Senatskanzlei, and the HTW Berlin University of Applied Sciences.
Furthermore we would like to thank Daniel Girardeau-Montaut (Cloudcompare) for his great
contribution to the 3D-community.
References
1. Boochs, F., Huxhagen, U., Kraus, K.: Potential of high-precision measuring techniques for
the monitoring of surfaces from heritage objects. In: International Workshop In-situ
Monitoring of Monumental Surfaces, Florence (2008)
2. Bandiera, A., Alfonso, C., Auriemma, R.: Active and passive 3D imaging technologies applied
to waterlogged wooden artifacts from shipwrecks. Int. Arch. Photogrammetry, Remote Sens.
Spat. Inf. Sci. 40(5), 15–23 (2015). Piano di Sorrento
3. Hess, M., Korenberg, C., Ward, C., Robson, S., Entwistle, C.: Use of 3D laser scanning for
monitoring the dimensional stability of a Byzantine ivory panel. Stud. Conserv. 60(sup1),
126–133 (2015). Sharjah
4. Arbace, L., Sonnino, E., Callieri, M., Dellepiane, M., Fabbri, M., Idelson, A.I., Scopigno, R.:
Innovative uses of 3D digital technologies to assist the restoration of a fragmented terracotta
statue. J. Cult. Heritage 14(4), 332–345 (2013). Pisa
5. Scopigno, R., Cignoni, P., Callieri, M., Ganovelli, F., Impoco, G., Pingi, P., Ponchio, F.: Using
optically scanned 3D data in the restoration of Michelangelo’s David. In: Optical Metrology,
pp. 44–53. International Society for Optics and Photonics, Munich (2003)
6. Wachowiak, M.J., Karas, B.V.: 3D scanning and replication for museum and cultural heritage
applications. J. Am. Inst. Conserv. 48(2), 141–158 (2009)
7. Saragusti, I., Karasik, A., Sharon, I., Smilansky, U.: Quantitative analysis of shape attributes
based on contours and section profiles in artifact analysis. J. Archaeol. Sci. 32(6), 841–853
(2005). Israel
8. Levoy, M., Pulli, K., Curless, B., Rusinkiewicz, S., Koller, D., Pereira, L., Ginzton, M.,
Anderson, S., Davis, J., Ginsberg, J., Shade, J., Fulk, D.: The digital Michelangelo project:
3D scanning of large statues. In: Proceedings of the 27th Annual Conference on Computer
Graphics and Interactive Techniques (Siggraph 2000), pp. 131–144. ACM Press/Addison-
Wesley Publishing Co., New York (2000)
9. Schäfer, A., Mara, H., Freudenreich, J., Bathow, C., Breuckmann, B., Bock, H.G.: Large scale
Angkor style reliefs: high definition 3D acquisition and improved visualization using local
feature estimation. In: Proceedings of the 39th Conference in Computer Applications and
Quantitative Methods in Archaeology, pp. 70–80 (2011)
10. Pitzalis, D., Cignoni, P., Menu, M., Aitken, G.: 3D enhanced model from multiple data sources
for the analysis of the cylinder seal of Ibni-Sharrum. In: VAST 2008: The 9th International
Symposium on Virtual Reality, Archaeology, and Cultural Heritage, pp. 79–84. Eurographics
Association, Braga (2008)
11. Fisseler, D., Weichert, F., Müller, G.G.W., Cammarosano, M.: Towards an interactive and
automated script feature analysis of 3D scanned cuneiform tablets. In: Scientific Computing
and Cultural Heritage, p. 16. Heidelberg (2013)
12. Zheng, S.Y., Huang, R.Y., Li, J., Wang, Z.: Reassembling 3D thin fragments of unknown
geometry in cultural heritage. ISPRS Ann. Photogrammetry, Remote Sens. Spat. Inf. Sci.
2(5), 393–399 (2014). Riva del Garda
13. Funkhouser, T., Shin, H., Toler-Franklin, C., Castañeda, A.G., Brown, B., Dobkin, D.,
Rusinkiewicz, S., Weyrich, T.: Learning how to match fresco fragments. J. Comput. Cult.
Heritage (JOCCH) 4(2), 7–13 (2011). New York
14. Goren, A., Kohlmeyer, K., Bremer, T., Kai-Browne, A., Bebermeier, W., Öztürk, D., Öztürk,
S., Müller, T.: The virtual archaeology project - towards an interactive multi-scalar 3D
visualisation in computer game engines. In: Across Space and Time, Papers from the 41st
Conference on Computer Applications and Quantitative Methods in Archaeology, pp. 386–
400. A. Traviglia Amsterdam University Press, Amsterdam (2013)
15. Rizzi, A., Voltolini, F., Remondino, F., Girardi, S., Gonzo, L.: Optical measurement
techniques for the digital preservation, documentation and analysis of cultural heritage. In:
VIII Conference on Optical 3D Measurement Techniques, Zürich, pp. 16–24, vol. 2 (2007)
16. Hassani, F., Moser, M., Rampold, R., Wu, C.: Documentation of cultural heritage; techniques,
potentials, and constraints. Int. Arch. Photogrammetry, Remote Sens. Spat. Inf. Sci. 40(5),
207–214 (2015). Göttingen
17. Gasparovic, M., Malaric, I.: Increase of readability and accuracy of 3D models using fusion
of close range photogrammetry and laser scanning. ISPRS - Int. Arch. Photogrammetry,
Remote Sens. Spat. Inf. Sci. 1, 93–98 (2012). Melbourne
18. Boehler, W., Marbs, A.: 3D scanning instruments. In: Proceedings of the CIPA WG, vol. 6,
pp. 9–18 (2002)
19. Iuliano, L., Minetola, P.: Rapid manufacturing of sculptures replicas: a comparison between
3D optical scanners. In: CIPA XX International Symposium, Turyn, pp. 384–389 (2005)
20. D’Apuzzo, N.: Overview of 3D surface digitization technologies in Europe. In: Proceedings
of SPIE, the International Society for Optical Engineering, pp. 1–13. Society of Photo-Optical
Instrumentation Engineers, San Jose (2006)
21. AICON 3D Systems breuckmann smartSCAN 3D- HE R5/C5 datasheet
22. Bernardini, F., Rushmeier, H.: The 3D model acquisition pipeline. Comput. Graph. Forum
21(2), 149–172 (2002). Blackwell Publishers Ltd.
23. Farouk, M., El-Rifai, I., El-Tayar, S., El-Shishiny, H., Hosny, M., El-Rayes, M., Gomes, J.,
Giordano, F., Rushmeier, H., Bernardini, F., Magerlein, K.A.: Scanning and processing 3D
objects for web display. In: 3DIM, Cairo, pp. 310–317 (2003)
24. Tucci, G., Cini, D., Nobile, A.: Effective 3D digitization of archaeological artifacts for
interactive virtual museum. In: Proceedings of the 4th ISPRS International Workshop 3D-
ARCH 2011, Florence (2011)
25. Besl, P.J., McKay, N.D.: Method for registration of 3-D shapes. In: Robotics-DL tentative,
pp. 586–606. International Society for Optics and Photonics, Boston (1992)
26. AICON 3D Systems OPTOCAT Version 2016R1 User Manual
27. Mautz, R., Tilch, S.: Survey of optical indoor positioning systems. In: 2011 International
Conference on Indoor Positioning and Indoor Navigation (IPIN), pp. 1–7. IEEE, Zürich (2011)
28. Barone, S., Paoli, A., Razionale, A.V.: Shape measurement by a multi-view methodology
based on the remote tracking of a 3D optical scanner. Opt. Lasers Eng. 50(3), 380–390 (2012)
29. UNIVERSAL ROBOTS UR10/CB3 Original instructions
30. CloudCompare (version 2.7) [GPL software] (2016). http://www.cloudcompare.org
PHOTOCONSORTIUM: Digitizing Europe’s
Photographic Heritage
Frederik Truyen1 ✉ and Antonella Fresa2

( )
1
KU Leuven, Leuven, Belgium
[email protected]
2
Promoter, Pisa, Italy
[email protected]
Abstract. Photoconsortium is an association of photographic archives that

contributed over 450.000 images of early photography to Europeana. In this
contribution we discuss lessons learned, in particular on digitization and copy‐
right issues and describe the activities involved in managing state-of-the-art digi‐
tized photographic archives. We discuss follow-on project activities such as
Europeana Space, which focuses on creative reuse of digitized cultural heritage
and the Europeana thematic photography channel.
Keywords: Photography · Digitisation · Europeana · Heritage · Cultural

heritage · Metadata · Archives
1 Introduction
EuropeanaPhotography [1] was an EC funded project under the ICT Policy Support
Programme of the Competitiveness and Innovation framework Programme, led by KU
Leuven and Promoter, which started in 2012 and successfully completed its work in
January 2015, delivering all the expected results.
The main outcome of the project was the delivery of hundreds of thousands of digi‐
tised masterpieces of early photography, coming from the most prestigious photographic
archives and museums in Europe. These digital collections were ingested into Europeana
[2], the European portal for digital cultural heritage, and are now accessible to
researchers, students, and all the interested people.
The partners of EuropeanaPhotography decided wisely to give a “new life” to the
outcomes of the project, and to make them alive beyond the end of the EC funded period.
The partners agreed on a long-term strategy, transforming cooperation synergies into
PHOTOCONSORTIUM Association, the International Consortium for Photo‐
graphic Heritage [3].
PHOTOCONSORTIUM association was established on the 29th of October 2014
and its first General Assembly took place in Bratislava on the 6th of November 2014
where President Prof. Fred Truyen and Vice-president Dr. Antonella Fresa were elected
together with the nomination of the members of the Steering Committee and of the
Executive Council. The aim of the association is to expand and enlarge the results of

DOI: 10.1007/978-3-319-48496-9_33
410 F. Truyen and A. Fresa
EuropeanaPhotography, providing a sustainability model to ensure the long-term

maintenance of the project results and to continue to cope with its original aims. In what
follows, we describe the work of PHOTOCONSORTIUM, during and after the Euro‐
peanaPhotography project.
2 Digitization of Photographic Cultural Heritage: Lessons Learned
Digitization of photographic cultural heritage involves many aspects, which we will

shortly discuss in what follows. First of all, photographic archives contain a diversity
of materials. Photographs can be kept as negatives, or as positive prints. The negatives
can be individual glass negatives, but also paper negatives (e.g. in the calotype process),
or they can be still part of celluloid film rolls, or cut out bands. Prints can be stand-alone
photographs, often in folders or in albums, or be part of published books. Many photo
houses also keep contact prints. The works in question could have been published or
not. Often archives hold unselected materials relating to famous published photographs
from famous photographers. Contact sheets are often revealing, as they show the selec‐
tion process, in many cases not only performed by the photographer himself but also by
the editor or publisher.
Managing a photographic archive hence involves a multitude of tasks, which have
to do with collection, documentation (metadata), curation, publication and preser‐
vation. In all these tasks, digitization and digital technologies play a major role. These
efforts need to keep in mind the key roles of an archive: protecting the authenticity and
integrity of the stored documents, preserving this cultural heritage and on the other hand
offering the public the means to exert their right to access this heritage. In Europeana‐
Photography, we gained several key experiences that we aim to provide as expert
knowledge to other photographic archives.
3 Collection
In EuropeanaPhotography, we set up a content committee to oversee what collections

were actually relevant to fit our goal, namely documenting the first one hundred years
of photography. We had several kinds of archives: the world famous collections of
Fratelli Alinari, Parisienne de Photographie, Imagno, Polfoto and TopFoto on the one
hand, which are funded privately, and public archives such as Arbeijdersmuseet from
Copenhagen, Denmark, ICCU/SGI in Italy or Gencat in Catalunya, Spain. Most of the
masterpieces in these collections are widely known. We also had an important group of
Central and Eastern European Archives which brought absolutely undiscovered material
to the table. Archives such as Divadelny Ustav from Slovakia, the Museum of History
of Photography in Krakow and ICIMSS from Poland, NALIS from Bulgaria, Lithuanian
Art Museums from Lithuania brought in unseen collections. In the case of ICIMSS and
NALIS, new material was researched and acquired from private family holdings. In
Germany, United Archives specializes in acquiring unknown photographic collections
to sell those to archives. Contributions were made by institutions that did not belong to
the original partnership, such as the Cyprus University of Technology, the specialized
PHOTOCONSORTIUM: Digitizing Europe’s Photographic Heritage 411
centre Bali LTD, the Soviéta Géografica Italiana and the Israel Museum. The National
Technical University of Athens provided technical support and the Koninklijke Musea
voor Kunst en Geschiedenis in Belgium offered metadata expertise. The Centre for
Image Research and Diffusion in Girona and the KU Leuven Imaging Lab took care of
digitization research.
4 Documentation and Metadata
In any of these different cases, it is important to know that large parts of current archives
are in fact undocumented, laying often unprocessed in maps and folders in their original
packaging when they came from the donations. In large part, the activity in European‐
aPhotography thus resulted in actually discovering part of the reality of the first one
hundred years of photography as it emerged from the archives. To have some consistency
in the descriptions and to allow thematic organization we setup a metadata committee
that developed a multilingual thesaurus in originally 13 – know 16 – languages
describing the techniques, styles and themes of early photography. This thesaurus has
been published as linked open data in skossified format [4]. Of course we are open to
accept translations in other languages and further additions to the vocabulary. It was the
first time that European photographic heritage was described using a common vocabu‐
lary. Since many of the established archives already used their own database and meta‐
data system, this required a mapping which we did using an intermediate LIDO standard,
which was the mapped to the Europeana Data Model using the widely used MINT tool
developed at NTUA. The metadata standard committee oversees the consistent use and
quality of the applied metadata. Besides correct photographic metadata we want good
content descriptions allowing for clear identification of the author, subject, place and
time of the photograph, which often poses serious problems as many photographs are
poorly documented by the original donators. This is BTW one of the compelling reasons,
as we will highlight further on, to publish these photographs in large databases such as
Europeana, as background metadata enrichment can take place derived from similar
photographs from other contributors. Problems with exact attribution of the author and
lack of exact date information also brings some copyright issues as we will discuss.
5 Curation
As photographic archives contain a wide diversity of materials, we implemented a two-

phased approach on curation. At first, each archive would look into its holdings and look
what themes emerged from the yet to be digitized material. This was then brought
together when partners showcased their collections to each other. This allowed us in a
second phase to determine topics of interest that would be pursued further, so that there
were transversal links between the partner collections. We decided e.g. not to focus too
much on the first world war, as there was a specific Europeana project working on this.
Themes would include a.o. leisure, sport, politics, art, culture, news and media. To make
the large EuropeanaPhotography collection (currently at more than 450.000 images)
more understandable and “tangible”, a selection was made for a showcase exhibition,
“All our Yesterdays”, [5] focusing on the lives in Europe in that era, with subthemes
such as “The City Lives”, “The Art of the Portrait”, “Yesterday’s Children”, “Hardship
and Drama”, “The Eye of the Beholder”, “Joy & Leisure” and “A Brave New World”.
A special subset “The Photographer at work” emerged from the finding that in many
collections a kind of “selfies” avant-la-lettre were discovered. The focus on the average
city lives was new to many of the archives, who usually focused on the important polit‐
ical, artistic and cultural heritage when curating collections. Rather than the famous poet
of renowned prime minister we wanted to show people strolling the streets, enjoying a
fair, going about their daily business. We discovered some lesser, more locally or even
squarely unknown photographers that easily deserve a place among the world famous
ones like Maurice Branger, Henry Cartier-Bresson and John Topham, such as Polish
photographer Tadeusz Rząca or the unknown German photographer Karl Heinrich
Lämmel (Fig. 1).
Fig. 1. Genre scene in the village, Krakow area. Rząca Tadeusz (1868–1928) – Museum of
History of Photography, Poland
6 Publication
Digitization of cultural heritage allows for publication of this heritage in a multitude of

formats and channels, thereby enhancing greatly access by the public. Communities
have the cultural right to access their heritage, as has been advocated in important EC
funded research such as Riches. The European Commission has made this goal one of
its policies, hence its continuing support for Europeana, the portal to Europe’s cultural
heritage. This means there is some pressure on museums, libraries and archives to open
up their content to the public. While this is quite natural for museums and libraries, this
is not so evident for archives. The whole point of an archive is to protect the documents
it keeps: it is absolutely essential that the very same, untampered document deposited
at an archive can be retrieved unaltered decades later. This often means that archives
have to shield documents in their holdings from changing moods and political turmoil.
Many archives have been threatened in the past when revolutions or civil strife took
place. Changes such as the move from Baltic countries out of the Soviet influence have
an impact on the expectations the public has about what is being published out of the
archive, and what is kept behind closed doors. There should be no illusion that there is
ever a thing such as simply opening up the whole contents of an archive to the public.
Besides societal pressures and political interference, two major problems are faced by
archive professionals when envisaging publication of archive contents: copyrights and
privacy.
As for copyrights, this is one of the toughest areas in digitization. While many legal
systems have provisions for archives that allow to make copies for preservation ends –
such as the US copyright law §108 – the same holds not true for publication. In particular,
this is an issue for publication on Europeana. This is the very reason why Europeana
developed a Licensing Framework, and provides in adapted Rights labels, partly based
on Creative Commons Licenses. Anyway, it is a quite laborious effort to determine
which works are in the Public Domain and which have rights attached to them that
prevent publication. EuropeanaPhotography built extensive expertise in this matter and
published useful tools and reports on this [6]. An even greater challenge is the so-called
“orphan works”, of which the author cannot be determined. Europe provides in an
Orphan Works Directive, to be implemented by the national authorities, to make it
possible for archives to publish these orphan works. Unfortunately, the procedure
remains very cumbersome and anyway involves a time-intensive “diligent search”,
making it impractical. Collecting societies strongly oppose this idea and want to charge
copyright indemnities, adding to the risk. Nevertheless, EuropeanaPhotography
published a large volume of reusable content in the form of Public Domain marked and
Creative Commons labelled works. In particular, readers might want to explore the open
collections of Generalitat de Catalunya and Lithuanian Art Museum (Fig. 2).
Privacy is another issue that needs careful handling. Many archives document
periods of strife and conflict. Certainly in smaller communities where families share
long histories opening up an archive to the general public might stir up emotions and
open old wounds. Serving the right of the public to access these testimonies of their
shared past requires a professional, respectful and balanced approach, as collections
themselves might be biased and do not necessarily reveal the whole story. It goes without
saying that, e.g., the opening of archives on the Spanish civil war is an issue of much
contention and debate.
Besides these two major issues other problems need some consideration. First of all
not all objects that are safely kept in an archive are fit for publication. Some are legally
forbidden, as is the case for some Nazi works, or would be perceived differently today
as they were in the past. As communities and ethics have diversified and evolved, many
documents or works that would have been perceived as neutral or harmless in past times
could be felt quite offensive or divisive today, for example as being utterly racist. A
related issue is the status of the original metadata. Depending on shifting ethics and
sensitivities, past documentation of works, e.g. the assigned title or description, could
be perceived today in a very different light. This is often the case with archives of colonial
heritage. This poses a dilemma: the public, in particular also stakeholders and
researchers, have a right to know the complete original metadata, while sometimes these
cannot be published within the confinements of the law. EuropeanaPhotography
members agreed in such cases to adapt the metadata, but to indicate that this has
Fig. 2. Portrait of the actress Míriam, dressed with Manila shawl and Cordovan hat. Antonio
Esplugas National Archive of Catalonia (ANC).
happened and that the original ones are available for consultation at the archive.
However, none of the partners has indicated that they actually had to use this procedure,
probably given the fact that our collections did not focus on conflict documentation.
A particular problem faced by EuropeanaPhotography is the concern of many private
contributors to archives about the moral integrity of the works or documents they
deposit. Many people would love to donate family photo heritage to their local archive,
but fear that when it becomes published online as public domain material, it could be
reused without any moral restraint. Photos of their family might be altered, “defaced”,
ridiculed and republished. Or worse: they could be used in computer shooting games!
The fear of desecration of personal or cultural heritage is important, and e.g. the Italian
law has a provision protecting Italian Art works against such misuse. In many cases
however, the owners think that only copyright protects them against such malpractice,
and so are reluctant to donate when these copyrights do no longer exist, as with public
domain works. However, in many jurisdictions there are solid protections of moral rights
that can be successfully defended in court, regardless of copyright. Anyway European‐
aPhotography was forced to develop rich documentation on these issues, and would
recommend archives to provide similar information to their stakeholder communities.
7 Preservation
Digitization has become part of any serious preservation strategy. Often because original
documents are prone to decay, and it is cheaper to preserve a digital copy. In the case
of photography, glass negatives and metal plate positives such as daguerreotypes or
tintypes preserve very well on their own, but the same certainly does not hold true for
paper and celluloid negatives. For the original period worked on by PHOTOCONSOR‐
TIUM, the first 100 years of photography, the main issue are the calotypes and albumen
prints, which are difficult to preserve and where digitization allows at least to keep the
information that we have now. However, since the focus of PHOTOCONSORTIUM
will move upwards in the 20th century, al lot of acetate and nitrate films in real danger
of being lost forever urgently need to be digitized.
This means of course that this digitization needs to be done at the highest standards.
For EuropeanaPhotography collections, with a large number of silver gelatine and wet
collodion glass plates, specific digitization procedures with backlighting and multiple
exposure in high dynamic range (HDR) were developed. For the KU Leuven collection,
e.g., a dual exposure method was developed so that both the glass diapositive photograph
as well as the surrounding frame with metadata could be captured in one image.
Current technology deployed amongst partners in PHOTOCONSORTIUM is very
well suited for glass and celluloid negatives and diapositives. For Daguerreotypes,
Ambrotypes and Tintypes research is being done with reflectance imaging techniques
to yield better results and capture the true properties of the analogue object (Fig. 3).
Fig. 3. Rogier Van der Weyden 1436 – Art History collection – KU Leuven Imaging Lab
The consortium decided on the best suited file formats to store and preserve the
photos, where more band more the Jpeg format comes into consideration as it has broad
industrial support. PHOTOCONSORTIUM established a liaison with the Jpeg standard
committee. In particular we are interested in encoding multispectral image data, in

fingerprinting and invisible watermarking, and interoperability issues. But PHOTO‐
CONSORTIUM also looks into the digital preservation cycle as a whole, on the long
term, and the physical storage and preservation. For this, common best practices were
developed and shared. They are available to the larger community through the PHOTO‐
CONSORTIUM website.
On the other hand maybe the best preservation of photographic cultural heritage is
to share it online, e.g. on Europeana. The more photographic testimonies are shared, the
better they survive. For repressive authorities, it is easier to shut down a physical archive
than to chase digital copies roaming the internet.
8 Results
The EuropeanaPhotography project produced excellent results. First of all, it digitised

and delivered to the European portal of cultural heritage Europeana about 450,000 of
the finest examples of early photography (1839–1939), increasing the value of the digi‐
tized photographic heritage retrievable via Europeana and accessible to everybody.
These photographs provide a unique insight into the evolution of European society and
the art of photography between the 19th and 20th centuries, which is a valuable source
for understanding our history from multiple approaches (history, economics, sociology,
anthropology, politics, art…). This extremely valuable source of information is
promoted by PHOTOCONSORTIUM through narratives able to attract the attention of
the general public and to contribute with novel information to the research. In this light,
PHOTOCONSORTIUM participates in the second edition of the Digital Service Infra‐
structure project coordinated by Europeana Foundation (DSI-2), with the task to develop
a photography channel in the Europeana portal. The new featured channel is planned to
become available online for public access in 2017.
The Consortium includes both public and private bodies with complementary back‐
grounds and it diffuse the useful guidelines about digitization and cataloguing of early
photography items which were produced in the European project. The Association also
gives access and maintains the multilingual vocabulary released by the project and
specifically tailored for photography, available in 16 languages1. PHOTOCONSOR‐
TIUM provides information IPR management and rights labelling of digital cultural
content, exploiting and re-using the extremely valuable contribution to a very lively
discussion which was carried out in the frame of the EuropeanaPhotography project. All
the knowledge resources produced by the project continue to be available and to be
maintained and updated by the PHOTOCONSORTIUM associates.
The project built a network of excellence on digitization and photographic heritage,
which continues to be active in the framework of the PHOTOCONSORTIUM Associ‐
ation. All the relevant results produced by the EuropeanaPhotography project are kept
updated in order to offer valuable solutions to the sector’s needs and to prepare to face
1
English, French, German, Dutch, Danish, Bulgarian, Slovak, Lithuanian, Polish, Spanish,
Catalan, Italian, Russian, Ukrainian, Chinese and Hebrew.
new challenges. The members of the PHOTOCONSORTIUM Association represent a

real aggregation of competences able to take care of updating technical documentation
and guidelines, taking into account the latest research and technical advances, and
promoting best practices and disseminating success stories. Cultural organisations,
photographic archives, research and educational institutions and any other party inter‐
ested in participating and contributing to the network can find in the PHOTOCON‐
SORTIUM Association the right place to meet experts and share experiences.
The purpose of the Association is the promotion and enhancement of the culture
of photography and the photographic heritage. To achieve its statutory goals,
PHOTOCONSORTIUM aims to promote, organize and manage conferences, exhibi‐
tions, awards and training courses as well as to carry out publishing activities. The
Association participates – and promote the participation of its members – in new projects
and initiatives, including but not limited to the participation in the Programmes of the
European Commission. All the initiatives of the association are primarily devoted to
serving the interests of its members. In this light, it carries out and encourages research
and studies on the themes of valorising and discovering photographic heritage, digitising
early photographic content, communicating and sharing photographic collections. The
association also continues to collect data and information of interest for the activities of
the members. Through the expertise of its members, the association provides consul‐
tancy and personalised services both to its members and to third parties.
9 Current Projects
Digital cultural heritage is considered by the EU one of the key drivers of economic
growth and social innovation. The Association is actively involved in European initia‐
tives and projects. PHOTOCONSORTIUM represents a centre of expertise and knowl‐
edge on digitization, aggregation of content to Europeana and other portals, metadata
standards, indexing, cataloguing and controlled vocabularies, best practices for the
management of digital archives, and much more. This expertise and knowledge is the
core to generate activities, provide services, organise training programs and seminars
and participate in new research.
The Association is the framework for participation in new initiatives and experi‐
mentation activities involving photographic heritage at the European level. As an
example, the digitized materials provided to Europeana through the EuropeanaPhotog‐
raphy project is the basis for the pilot activities in the Europeana Space [7] project.
Europeana Space - Spaces of possibility for the creative re-use of digital cultural content
- is a Best Practice Network project funded by the European Union under the ICT Policy
Support Programme, whose project coordinator is Coventry University and technical
coordinator is Promoter S.r.l. The network aims to increase and enhance the creative
industries’ use and re-use of Europeana and other digital collections. Multiple themed
pilots present innovative models of digital cultural content exploitation in interactive
TV, dance, games, publishing, museums, and, namely, photography.
PHOTOCONSORTIUM cooperates in the photography pilot of Europeana Space
under the coordination of KU Leuven.
The pilot on photography is experimenting the reuse of the content aggregated

through EuropeanaPhotography and demonstrating a range of possibilities offered by
apps, the Europeana API2, the Technical Platform of Europeana Space, and a multitude
of tools developed by the open source community, to come up with innovative models
involving photography. The experimentation activities include a hackathon (which took
place in Leuven in February 2016), where developers of novel museum applications,
storytelling and augmented reality apps met and proposed their projects. The best ideas
were tunnelled through a monetization event and currently incubated to prepare for
launch on the market in 2017.
At the same time, PHOTOCONSORTIUM Association cooperates with crowd‐
sourcing and citizen science initiatives. Photographic heritage is everywhere around us
and helps us to enlighten personal and community stories, by visually transmitting ways
of life and all kinds of experiences. The citizens’ involvement in photographic heritage
activities is a key element for the Association, not only as audience but also as an active
actor in the activities promoted by PHOTOCONSORTIUM. An example of how to
activate public participation in photographic heritage projects was the public digitisation
activities carried out in the framework of “All Our Yesterdays” [8], the important photo‐
graphic and multimedia travelling exhibition which showcases the best examples of the
photographs digitised in EuropeanaPhotography. The first round of this exhibition was
held in April–June 2014 in Pisa and the second in January–March 2016 in Leuven. Both
editions included as a side element a digitisation station: a dedicated space provided
with advanced digitisation technology and curated by professionals of photography,
where visitors of the exhibition brought their own vintage photos to be digitised. In fact,
a large part of our photographic cultural heritage is preserved and hidden in family
albums, which not only include wedding pictures and grandfathers’ portraits, but also
reveal photographs of great historical and societal interest. It is of the utmost importance
that citizens are encouraged to share this wealth of cultural heritage with the whole
community. To explore more in depth how to address this participatory research, the
association established a cooperation agreement with CIVIC EPISTEMOLOGIES [9],
a successful European project, coordinated by the Italian Ministry of Economic Devel‐
opment. The main outcome of CIVIC EPISTEMOLOGIES is the roadmap for Citizen
Researchers in the Age of Digital Culture. The collaboration between PHOTOCON‐
SORTIUM and CIVIC EPISTEMOLOGIES contributed to activate citizen involvement
in photographic heritage research.
The participation of PHTOCONSORTIUM in the Digital Service Infrastructure
project coordinated by Europeana Foundation is the most recent example of commitment
of the Association to participate in European initiatives. The project, which lasts from
July 2016 until beginning of 2018, foresees the creation of a range of thematic channels
to promote access, use and re-use of digital cultural heritage content. Within the DSI-2
project, PHOTOCONSOTIUM curates the creation of the thematic channel on photo‐
graphic heritage. This is a great acknowledgement of the value of Association within
the European framework of research and development, which is expected to continue
in the coming years through new initiatives.
2
Application Program Interface.
The Association’s commitment to photographic culture and heritage brings its

members to proactively lead initiatives to value and transmit this rich heritage. The
Association creates specific programs to reach all target audiences, including cultural
institutions, photographic archives, specific research communities, the educational
sector, younger and elder generations, the general public, and activists and amateurs in
general.
The promotion and enhancement of photographic heritage finds in the Association
a valuable tool to reach its objectives: a dedicated website giving visibility to all the
digitised materials. Complementary to PHOTOCONSORTIUM website, a dedicated
showcase was established in the digitalmeetsculture magazine edited by Promoter S.r.l.
(one of the founding members of PHOTOCONSORTIUM) which counts currently more
than 25,000 visits per month [10].
PHOTOCONSORTIUM was established as a non-profit association with a demo‐
cratic structure, political independence, and unlimited duration. PHOTOCONSOR‐
TIUM allows membership to both natural and legal persons, public or private bodies,
companies, and associations. It aims to enlarge its network internationally and welcomes
new members sharing its aims.
More details about PHOTOCONSORTIUM and its future activities are available
online at: www.photoconsortium.net and on our blog Digital Meets Culture.
Join our community by contacting [email protected].
Acknowledgements. EuropeanaPhotography funding grant European Union CIP-ICT-PSP-2011-

5/297158; Europeana Space funding grant European Union CIP-ICT-PSP-2013-7/621037; CIVIC
Epistemologies funding grant European Union FP7-INFRASTRUCTURES-2013-2/632694.
References
1. EuropeanaPhotography: http://www.europeana-photography.eu
2. Europeana: http://www.europeana.eu
3. PHOTOCONSORTIUM: http://www.photoconsortium.net
4. EuropeanaPhotography Multilingual Early Photography thesaurus: http://bib.arts.kuleuven.be/
photoVocabulary/en.html
5. Exhibition “All our Yesterdays”: http://www.earlyphotography.eu
6. Truyen, F., Waelde, C.: Copyright, cultural heritage and photography: a gordian knot? In:
Borowiecki, K., Forbes, N., Fresa, A. (eds.) Cultural Heritage in a Changing World, pp. 77–98.
Springer, Cham (2016). http://www.europeana-photography.eu/index.php?en/117/documents
7. Europeana Space: http://www.europeana-space.eu
8. http://www.earlyphotography.eu/
9. CIVIC Epistemologies: http://www.civic-epistemologies.eu/
10. Digital Meets Culture: http://www.digitalmeetsculture.net
Acquisition and Processing Experiences of Close Range
UAV Images for the 3D Modeling of Heritage Buildings
Arnadi Murtiyoso1, Pierre Grussenmeyer1 ✉ , Mathieu Koehl1, and Tristan Freville2

( )
1
Photogrammetry and Geomatics Group, ICube Laboratory UMR 7357, INSA Strasbourg,
24 Boulevard de la Victoire, 67084 Strasbourg, France
{arnadi.murtiyoso,pierre.grussenmeyer,
mathieu.koehl}@insa-strasbourg.fr
2
Drone Alsace, 4 Rue Sainte Cathérine, 67000 Strasbourg, France
[email protected]
Abstract. The use of image-based techniques to document heritage sites has

seen a resurgence in recent years with advancements in optical sensors as well as
computing power. The rise of UAVs (Unmanned Aerial Vehicles) also comple‐
ments this technique, by providing the advantage of aerial view over traditional
terrestrial image acquisition. Recently UAVs began to become more and more
specialized towards specific tasks, 3D modeling and reconstruction being some
of them. In this study the use of state of the art UAV dedicated for close range
inspection is analysed. Several case studies were performed on historical build‐
ings in Strasbourg, France. Processing was done by utilizing both commercial
and open source photogrammetry and SfM (Structure from Motion) solutions.
Both the quality of the aerotriangulation and the dense matching were studied.
The final objective of this project is to adapt existing terrestrial image acquisition
and processing protocols for use by UAVs.
Keywords: UAV · Documentation · Photogrammetry · Close range · Protocols ·

Heritage buildings
1 Introduction
Since its first conception as military support, the UAV has seen a significant shift towards
civilian use. In the fields of geomatics, this spread of small scale aerial technology means
that it could be used to perform close range aerial photogrammetry. Coupled with the
rapid development of optical sensors as well as computing power, this type of surveying
becomes a very potential solution for various uses. The field of heritage documentation
naturally benefits from these developments, as it complements the already established
method of image-based terrestrial techniques [1].
Several types of UAVs exist in the market nowadays. A general classification of
UAV types was given by [1, 2] which divides it into three main categories according to
their physical features:

DOI: 10.1007/978-3-319-48496-9_34
Acquisition and Processing Experiences of Close Range UAV Images 421
• Lighter than air platforms, such as balloons and kites. This category is low-cost but
is more difficult to control due to its low wind resistance and low velocity.
• Fixed-wing platforms, with the capability of covering a large surface but may be
limited in payload as well as wind resistance. The fixed-wing UAV is suited for
larger-scale mapping resembling classical small-format aerial photogrammetry.
• Rotary-wing platforms, either with a single or multiple rotors. This type of UAV has
a larger wind resistance, but its surface coverage can be significantly lower than that
of the fixed-wing type. Hence its vast use in close-range applications.
In the field of geomatics, the optics as well as the sensor of the on-board camera
plays an important role. Some UAV manufacturers have tried to accommodate these
geometric demands by integrating better quality lenses, although it is still often limited
by the payload. More recently, UAVs have started to specialize in various specific
sectors such as agricultural mapping and close industrial (and heritage) inspection.
Progress in the field of computer vision has largely facilitated the classical photo‐
grammetric workflow. As explained in [2, 3], the typical workflow involves automatic
tie point feature extraction and matching. This is often followed by robust outlier detec‐
tion and elimination and bundle block adjustment in order to retrieve the position and
orientation of each camera station. Afterwards, dense matching algorithms enable the
3D reconstruction of a dense point cloud, up to one point for each pixel of the image [4].
2 Related Works
UAVs have seen wide use in the documentation of archeological sites, often to provide
a general view of the surroundings [5]. This can be done either with a fixed-wing [6] or
rotary-wing UAV [5, 6] depending on the dimensions of the site in question. Another
use of the UAV is for close-range modeling or inspection of buildings by using the
rotary-wing type. This type of acquisition is often performed in conjunction with terres‐
trial images and sometimes also with range-based techniques such as terrestrial laser
scanning [7]. In these cases the UAV presents a natural advantage over other terrestrial
techniques in its capability to capture aerial images, thus covering angles which would
otherwise be impossible to cover from the ground.
In regards to acquisition and processing protocols, several procedures exist in the
literature which deals mainly with terrestrial images. Often these protocols deal not only
with the acquisition, but also with the calibration and recommendations for facilitating
later processing steps. One of the objectives of this study is to adapt the existing estab‐
lished terrestrial acquisition protocols for use by UAVs.
The CIPA Heritage Documentation’s 3 × 3 rules have existed since 1994 with several
updates [8]. The latest version can be consulted in the CIPA website1. These rules
concern not only the acquisition of images, but also recommendations for the calibration
and procedures to be implemented during the survey.
Another protocol called TAPENADE (Tools and Acquisition Protocols for
Enhancing Artifacts Documentation) was developed for the documentation of cultural
1
http://cipa.icomos.org/.
422 A. Murtiyoso et al.
heritage [9, 10]. These documents provide several different cases of objects and gener‐
ally follow the image acquisition configuration suitable for the software Micmac.
However, up until the writing of this article, only French versions of these protocols are
available although a more general explanation of the recommendations in English is
available in the TAPENADE website2.
The University of Stuttgart’s Institute for Photogrammetry has also developed their
own protocol [11]. This protocol, called “One panorama each step”, is targeted for
producing a good dense matching result by using greatly overlapping images. This
protocol was developed and tested for the dense matching software SURE and has also
been tested on UAV data.
These existing protocols share some similarities. First, a good calibration using a
convergent, controlled environment is always recommended rather than relying solely
on in situ self-calibration. The importance of procedural documentation and description
of the object is also stressed, something which may easily be forgotten but is nevertheless
important. Finally, all three rules emphasize two kinds of image acquisition. The first
involves general or global images with good convergence angles to ensure the precision
of the resulting network while the second involves images with very large percentage
of overlap (detailed stereo pairs) to facilitate the dense matching process.
3 UAVs and Software Solutions Employed
The first UAV used in this project is a multi-copter rotary wing platform manufactured
by the company Sensefly and called Albris since April 2016 (previously known as
Exom). The product first appeared in the market in May 2015 and was aimed to perform
close-range high resolution inspections, mapping, and 3D modeling. It is equipped with
multiple sensors, including a still 38 megapixels camera, a thermal and a video camera.
Several ultrasonic sensors give an approximate distance measurement of its surround‐
ings, enabling it to fly on a set distance from an object. The still camera itself is furnished
with an 8 mm lens and a 10 × 7.5 mm sensor. This specification theoretically enables a
GSD (Ground Sampling Distance) of up to 1 mm at a distance of 6 m.
Another UAV used is the DJI Phantom 3 Professional, also a multi-copter rotary
wing type. This UAV was first released also in May 2015. The Phantom 3 is lightweight
and relatively low-cost, but it is not geared for dedicated close-range inspection tasks.
Unlike the previous versions, the Phantom 3 is no longer equipped with a fish-eye lens
camera. This could potentially yield a better geometric result, but the sensor itself
remains very small compared to most terrestrial cameras. It is equipped with a 4 mm
lens and a 6.5 × 5 mm sensor. Figure 1 shows the main characteristics of these two
UAVs.
The software solutions used in this study ranges from the commercial to the open
source. From the commercial side Pix4D, Agisoft Photoscan and Photomodeler are used.
Meanwhile open source solutions such as Apero-Micmac, SURE, and VisualSFM/
PMVS2 were tried.
2
http://www.tapenade.gamsau.archi.fr/.
Sensefly Albris
Platform
Payload 1.8 kg
Flight autonomy ~22 minutes
Camera
Focal length 8 mm
10 x 7.5 mm
Sensor size
7152 x 5368 pixels
Horizontal FOV 630
DJI Phantom 3 Professional

Platform
Payload 1.3 kg
Flight autonomy ~23 minutes
Camera
Focal length 4 mm
6.5 x 5 mm
Sensor size
4000 x 3000 pixels
Horizontal FOV 940
Fig. 1. The UAV used in the study, Sensefly Albris and DJI Phantom 3 Professional as well as
the main characteristics of their respective platforms and on board cameras.
From an algorithmic point of view, several approaches to the dense matching

problem exist. The article [12] explains well the different types of dense matching algo‐
rithms. These different approaches towards dense matching may give different results
in different scenarios.
Pix4D and Photoscan are two commercial solutions with a rather black-box nature.
However, concurrent with the results of [12], Photoscan most probably performs a
modified SGM (Semi-Global Matching) [13] of stereo pairs to generate depth maps, and
then employs epipolar constraints at the end of this process to filter the results. Pix4D
may have used a similar albeit different approach, since an SGM-based matching is
offered as an additional plug-in.
Photomodeler has long been used by the architecture and archeology community for
performing 3D image-based measurements [8]. This software has an advantage over
other commercial solutions in that it provides more statistical information on its results,
making it less black-box like. Photomodeler has since added a dense matching module
based on stereo-pairs and more recently on a multi-view geometry.
The approach taken by SURE is well described in [14]. This software also uses a
modified version of the SGM and computes a disparity map for each potential pair of
images with a set value of overlap. The resulting depth-maps are then converted into a
point cloud by employing geometric constraints to help reduce the number of outliers
[12]. A free version of SURE is available for research or academic purposes.
Meanwhile, Micmac uses a multi-resolution and multi-image approach to dense
matching [15]. The Micmac suite is modular and works with several levels of complexity
and automation. The first module, Pastis, searches and matches tie points on the images.
Apero performs bundle block adjustment to retrieve the external orientation parameters
of the camera stations. Finally, Micmac performs a pyramidal processing to search pixel
correspondences. Results from a lower resolution matching are used to guide the
matching at the higher resolution level, with the maximum resolution determined by a
parameter [12].
PMVS (patch-based multi-view stereo) uses a different object-based approach [16].
It implements multi-view stereo-matching starting on a sparse point cloud generated by
the SfM matched tie points. The matched “patch” around a tie point is then repeatedly
expanded to nearby pixels and filtered using visibility constraints.
4 Case Studies
The UAVs were deployed to acquire images of two historical buildings located in the
city of Strasbourg, France. The choice of objects was determined by their particularities
and therefore also challenges, e.g. façades, painted walls, columns, etc.
4.1 The Josephine Pavilion
The Josephine Pavilion is a 19th century building located inside the Orangerie Park in
the city of Strasbourg. It provides a good test object as the area around it is open without
much vegetation. The back-façade is made from red sandstone, typical of the Alsace
region while the front-façade is a relatively smooth wall painted in beige, giving little
texture for the dense image matching step.
A test flight was performed in late 2014 before the project started using the Phantom
3 UAV. Close range images of each of the four façades from an average distance of 8 m
were taken. A quick processing of these images (Fig. 2) shows dense matching problems
on the front-façade, which has very little texture. A comparison of dense matching results
of the central part of this façade using the employed matching algorithms is shown in
Fig. 3.
Fig. 2. The textured mesh of the Josephine Pavillion data set generated by Photoscan. Note the
meshing problem at the right side of the central façade due to lack of points.
Fig. 3. Dense matching results of the central front-façade of the Josephine Pavilion, showing
point clouds generated by the five dense matching algorithms employed.
Visually, Photoscan managed to retrieve points on the problematic beige wall.

However, these points are noisy with up to 15 cm of dispersion on a wall which normally
should be flat. Similarly with Photoscan, SURE also generates points on the painted
surface with noises. On the contrary, processing on Photomodeler (using its MVS dense
matching mode), Micmac, and PMVS detected only a sparse amount of points on this
smooth surface.
Several preliminary analyses were taken from this pilot project. The first concerns
the importance of image overlap of the object; indeed echoing the need for detailed
images with large percentage of overlap mentioned in the existing protocols. A remark
was made to acquire controlled oblique images in further projects in order not only to
strengthen the image network geometry but also to better cover hidden parts (window
sills, awnings, balcony, etc.). Oblique images will emulate the concept of detailed stereo
pairs in a UAV acquisition scenario.
Furthermore, the problem of the object in question’s material is also important to
note. In the case of the pavilion, a textureless material hinders a proper pixel-by-pixel
search for correspondence. Particular strategies should be taken to address this problem
[12]. One strategy that can be employed involves changing the resolution of the images
used in the dense matching; bearing in mind that in this case a better resolution matching
does not necessarily produces a better result. Another involves changing the correlation
coefficient threshold to allow more points (which may however be noisy) to be detected.
4.2 The Rohan Palace
The Rohan Palace is a historical landmark of the city of Strasbourg dating to the 18th
century. Located next to the cathedral, the palace was built for the Cardinal Rohan
between 1732 and 1742 and housed several French sovereigns during their visits to
Strasbourg. Today it is used by three museums, all managed by the city’s administration.
In regards to the material, the palace is made of yellow sandstone. For this project, only
the central façade overlooking the River Ill was photographed. The dimension of this
façade is approximately 14 × 20 m.
As decided and based on previous experiences on the Josephine pavilion, a classical
perpendicular flight was performed followed by four oblique flights with the cameras
tilted to the right, left, up, and down (Fig. 4). The configuration of flight strips is kept
the same for all five flights in order to emulate a highly overlapping and convergent
geometry and thus cover difficult angles and partly hidden parts. The disadvantage of
this method, however, is that the same flight plan must be performed five times. This
increases flight time and therefore also poses problem in regards to battery issues and
image texture homogeneity.
Fig. 4. Perpendicular and oblique flight plans used for the façade.
In this project, the Sensefly Albris was used to acquire the images. Using its capa‐
bility to fix camera-object distances, the UAV was flown at a fixed approximate distance
of 5 m from the façade. In addition, 13 control and check points scattered on the façade
are measured using a total station using the spatial intersection method in order to be
able assess the precision and accuracy of the results. The 6 control points are placed on
the limits of the flight zone following the classical aerial photogrammetry configuration.
The remaining 7 points were used as check points, and are scattered evenly on the façade
to represent changes of reliefs.
All algorithms succeeded in orienting the images except VisualSFM, which might
be linked to the default number of detected tie points. Despite the theoretical GSD of
1 mm, here the precision of the aerotriangulation for all four solutions was on average
9 mm while the check point accuracy was 7 mm. It should be noted that the control and
check points used are detail points (window edges, bricks, etc.) and not well defined
targets. The precision and accuracy depend therefore on the quality of point marking.
This however, was hampered by the noisy quality of the Albris’ images. This noise
problem has been acknowledged by Sensefly and an improvement of sensor quality is
expected sometime soon.
As regards to the dense point cloud, the results as well as matching parameters
employed for the four dense matching algorithms are shown by Fig. 5. The matching
settings used in each algorithm correspond to the resampling of the input images to a
quarter of their original resolutions. Note that this resolution setting is often employed
in dense matching solutions in order to give the users a compromise between quality
and processing time. The application of oblique images has effectively covered several
difficult places such as the balcony and the lower part of the triglyph or the metope above
it. However, the question of image noise still posed a problem. Indeed, a test performed
with matching using half resolution images generated a sparser point cloud with many
holes. On the contrary, a more complete point cloud was acquired using a lower reso‐
lution setting. This problem is most probably caused once again by the correlation
coefficient threshold; in the higher resolution setting the algorithm calculates denser
points which it assumes are noises and therefore deletes, leaving holes in the resulting
point cloud.
Fig. 5. Dense point cloud generation parameters of the four algorithms tested.
Furthermore, in order to validate the accuracy of the dense matching results, a

terrestrial laser scanning survey was performed on the façade concerned. This was done
using a Faro Focus 3D laser scanner and the resulting point cloud has a resolution of
about 6 to 8 mm. Evaluation was performed for a common portion of each result of the
photogrammetric dense matching using the laser scanner point cloud as reference (see
Fig. 6). All solutions gave standard deviations of around 1 cm compared to the reference.
This corresponds more or less to the theoretical resolution of the point cloud at this image
pyramidal level of matching (around 1.4 cm).
Larger histogram dispersion is observed in Photomodeler’s results, which indicates
a noisier point cloud. More holes are also observed in this point cloud. Photoscan and
Pix4D gave more homogeneous i.e. less noisy results, with Photoscan detecting slightly
more error. In addition, Micmac detects more holes in general compared to Pix4D and
Photoscan, which may be related to the default correlation coefficient threshold in its
semi-automatic C3DC matching mode.
In Fig. 7, the cross-section of one of the Corinthian columns is analyzed. The point
cloud generated by Photomodeler presents the most noise. Micmac’s results present a
Photoscan, σ =0.011 m Pix4D, σ = 0.010 m
Micmac, σ = 0.011 m Photomodeler, σ = 0.013 m
Fig. 6. Dense matching results analysis using laser scanning data as reference for a common part
of the façade of the Rohan Palace.
circular trend for the cross-section albeit it is a rather rough, unsmooth representation.
Photoscan and to some degree also Pix4D, present an almost smooth circular shaped
point cloud profile. This suggests a form of interpolation and/or smoothing performed
after the matching process to conform to certain geometric constraints. Furthermore,
some solutions had difficulties in detecting points at the junctions where the column and
the wall meet. Micmac, Photomodeler, and even the laser scanner (at the right junction)
Laser
Photoscan σ = 6.7 mm
Pix4D σ = 5.4 mm
Micmac σ = 4.5 mm
Photomodeler σ = 10.2 mm
Horizontal Vertical
Profile Profile
Laser Pix4D σ = 6.6 mm Photomodeler σ = 8.6 mm

Photoscan σ = 7.0 mm Micmac σ = 6.9 mm
Fig. 7. A horizontal profile of one of the Corinthian columns (red square) and a vertical one of
a part of the wall (green square). The blue lines represent laser scanner measurements. (Color
figure online)
failed to detect points in this zone and therefore leave a hole. It is however interesting
to note that both Photoscan and Pix4D still manage to detect points and close the hole
in these difficult areas, although at the left junction Photoscan seems to have detected
false points behind the column. Micmac and Photomodeler were also unable to recon‐
struct the left junction. However, Photoscan and Photomodeler are both less accurate in
this case compared to Micmac and Pix4D.
Horizontal Profile Error Vertical Profile Error
Photoscan Photoscan
Pix4D Pix4D
Micmac Micmac
Photomodeler Photomodeler
Fig. 8. Error graphs for the horizontal and vertical profiles using the same profiles from the laser
data as reference.
Figure 7 also describes the vertical profile of a portion of a wall generated by the
different algorithms. Again, results from Photoscan present a continuous surface which
seems to be smoothed to some degree. This phenomenon is however less evident in the
Pix4D profile although it can still be observed in a smaller scale. In this regard Pix4D
is also a little noisier than Photoscan. Similar to the previous case, Micmac gives a result
which follows the trend of the reference with minor noises. Photomodeler gives the
noisiest point cloud with errors of up to 1 cm (Fig. 8).
5 Conclusions and Further Work
In this paper, the use of UAVs for the documentation of heritage buildings was assessed.
In regards to the acquisition, oblique flights were performed to cover difficult angles and
to emulate the detailed stereo pairs taken in terrestrial close range photogrammetry
surveys. These oblique flights give a much larger overlap for each point of the object
and therefore enable the dense matching algorithm to generate a denser, more complete
point cloud.
The processing of these UAV images stress the importance of object textures despite
state of the art pixel by pixel matching. It also stresses on the quality of the basic input,
which is the images themselves. Even with more pixels in an image, image noises can
render them useless and therefore reduce the quality of the matching. The comparison
of several software solutions has also been addressed in this article. Even though all
solutions give similar overall accuracies, the matching algorithms give differing results.
A closer analysis showed that some algorithms may have added a point cloud post-
processing step at the end of the matching.
In conclusion, the UAV has shown to be very potential to perform close range
heritage inspections, being able to reach a centimetric (and theoretically even less)
accuracy. It’s capability to capture aerial images is a clear advantage compared to
terrestrial techniques, and the use of oblique images in this study has been proven to be
effective in covering otherwise difficult parts of the object. The software solution chosen
should follow the cases encountered as they each have their own advantages and disad‐
vantages. An acquisition of a more complicated building will be performed during this
project in order to implement these lessons and experiences.
References
1. Nex, F., Remondino, F.: UAV: platforms, regulations, data acquisition and processing. In:
Remondino, F., Campana, S. (eds.) BAR International Series, vol. 2598, pp. 73–86.
Archaeopress, Oxford (2014)
2. Remondino, F., Pizzo, S., Kersten, Thomas, P., Troisi, S.: Low-cost and open-source solutions
for automated image orientation – a critical overview. In: Ioannides, M., Fritsch, D., Leissner,
J., Davies, R., Remondino, F., Caffo, R. (eds.) EuroMed 2012. LNCS, vol. 7616, pp. 40–54.
3. Chiabrando, F., Donadio, E., Rinaudo, F.: SfM for orthophoto generation: a winning approach
for cultural heritage knowledge. In: International Archives of the Photogrammetry, Remote
Sensing and Spatial Information Sciences, vol. XL-5/W7, pp. 91–98 (2015)
4. Achille, C., Adami, A., Chiarini, S., Cremonesi, S., Fassi, F., Fregonese, L., Taffurelli, L.:
UAV-based photogrammetry and integrated technologies for architectural applications –
methodological strategies for the after-quake survey of vertical structures in Mantua (Italy).
Sensors 15, 15520–15539 (2015)
5. Remondino, F., Barazzetti, L., Nex, F., Scaioni, M., Sarazzi, D.: UAV photogrammetry for
mapping and 3D modeling – current status and future perspectives. In: International Archives
of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XXXVIII-1/
C22, pp. 25–31 (2011)
6. Suwardhi, D., Menna, F., Remondino, F., Hanke, K., Akmalia, R.: Digital 3D Borobudur:
integration of 3D surveying and modeling techniques. In: International Archives of the
423 (2015)
7. Grenzdöffer, G., Naumann, M., Niemeyer, F., Frank A.: Symbiosis of UAS photogrammetry
and TLS for surveying and 3D modeling of cultural heritage monuments – a case study about
the cathedral of St. Nicholas in the City of Greifswald. In: International Archives of the
96 (2015)
8. Grussenmeyer, P., Hanke, K., Streilein, A.: Architectural photogrammetry: basic theory,
procedures, tools. In: Kasser, M., Egels, Y. (eds.) Digital Photogrammetry, pp. 300–339.
Taylor & Francis, London (2002)
9. Pierrot-Deseilligny, M., De Luca, L., Remondino, F.: Automated image-based procedures for
accurate artifacts 3D modeling and orthoimage generation. In: Proceedings of the XXIIIrd
International CIPA Symposium (2011)
10. Nony, N., Luca, L., Godet, A., Pierrot-Deseilligny, M., Remondino, F., Dongen, A., Vincitore,
M.: Protocols and assisted tools for effective image-based modeling of architectural elements.
In: Ioannides, M., Fritsch, D., Leissner, J., Davies, R., Remondino, F., Caffo, R. (eds.)
EuroMed 2012. LNCS, vol. 7616, pp. 432–439. Springer, Heidelberg (2012). doi:
10.1007/978-3-642-34234-9_44
11. Wenzel, K., Rothermel, M., Fritsch, D., Haala, N.: Image acquisition and model selection for
multi-view stereo. In: International Archives of the Photogrammetry, Remote Sensing and
Spatial Information Sciences, vol. XL-5/W1, pp. 251–258 (2013)
12. Remondino, F., Spera, M.G., Nocerino, E., Menna, F., Nex, F.: State of the art in high density
image matching. Photogram. Rec. 29(146), 144–166 (2014)
13. Hirschmûller, H.: Semi-global matching – motivation, developments and applications. In:
Photogrammetric Week 2011, pp. 173–184 (2011)
14. Wenzel, K., Rothermel, M., Haala, N., Fritsch, D.: SURE – the ifp software for dense image
matching. In: Photogrammetric Week 2013, pp. 59–70 (2013)
15. Pierrot-Deseilligny, M., Paparoditis, N.: A Multiresolution and optimization-based image
matching approach: an application to surface reconstruction from SPOT5-HRS stereo
imagery. In: International Archives of the Photogrammetry, Remote Sensing and Spatial
Information Sciences, vol. XXXVI-1/W41 (2006)
16. Furukawa, Y., Ponce, J.: Accurate, dense, and robust multi-view stereopsis. IEEE Trans.
Pattern Anal. Mach. Intell. 32(8), 1362–1376 (2009)
Internal 3D Printing of Intricate Structures
Théophane Nicolas1, Ronan Gaugne2, Cédric Tavernier3,

Valérie Gouranton4 ✉ , and Bruno Arnaldi4
( )
1
Inrap/UMR 8215 Trajectoires, Cesson-Sévigné, France
2
Université de Rennes 1/IRISA-Inria, Rennes, France
3
Image ET, Mordelles, France
4
INSA Rennes/IRISA-Inria, Rennes, France
[email protected]
Abstract. Additive technologies are increasingly used in Cultural Heritage

process, for example in order to reproduce, complete, study or exhibit artefacts.
3D copies are based on digitization techniques such as laser scan or photogram‐
metry. In this case, the 3d copy remains limited to the external surface of objects.
Medical images based digitization such as MRI or CT scan are also increasingly
used in CH as they provide information on the internal structure of archaeological
material. Different previous works illustrated the interest of combining 3D
printing and CT scan in order to extract concealed artefacts from larger archaeo‐
logical material. The method was based on 3D segmentation techniques within
volume data obtained by CT scan to isolate nested objects. This approach was
useful to perform a digital extraction, but in some case it is also interesting to
observe the internal spatial organization of an intricate object in order to under‐
stand its production process. We propose a method for the representation of a
complex internal structure based on a combination of CT scan and emerging 3D
printing techniques mixing colored and transparent parts. This method was
successfully applied to visualize the interior of a funeral urn and is currently
applied on a set of tools agglomerated in a gangue of corrosion.
Keywords: Archaeology · Tomography · 3D printing
1 Introduction
Cultural Heritage professionals such as archaeologists and conservators regularly expe‐

rience the problem of working on concealed artefacts and face the potential destruction
of source material without real understanding of internal structure or state of decay. For
example, artefacts may be encased in corroded materials or a block of ash, or integrated
with, and inseparable from, larger assemblies, such as manufactured objects composed
of several pieces.
The work presented in this paper aims to develop a non-destructive workflow for
analysing and documenting the internal structure of artefacts, improving diagnostic
techniques and knowledge by combining medical imaging technologies such as
Computed Tomography (CT) with 3D printing. Our methodology allows to create digital

DOI: 10.1007/978-3-319-48496-9_35
Internal 3D Printing of Intricate Structures 433
and physical 3D surrogates of objects to investigate, analyse and interpret their internal
structure through volumetric scanning, 3D image rendering, and 3D printing.
Combinations of CT scan and 3D printing technologies have already been proposed
in CH contexts. The projects presented in [1] and [2] both propose the use of 3D printers,
combined with CT images, to reproduce high-value pieces, Chinese chess pieces in the
first case, and Gold jewels in the second case. In [10], copies of fragile bones are used
for sharing during study process and for exhibition.
This methodology was also successfully applied to physically access to encased
artefacts through 3D images, and 3D printed replications without any irreversible phys‐
ical action on the original material [3], even in the case of a disaggregated artefact [4].
In particular, in the first work, a removable copy of a Gallic weight illustrates how this
technique allowed to better understand the internal structure of an artefact. This is also
the case in the work of McKnight et al. [5] where a bones assembly from an animal
mummy has been reconstituted for scientific analysis and public exhibition purpose. In
both [2] and [6], a 3D print of the original shape of the archaeological material composed
of an aggregate of several objects is proposed. This kind of production is useful to study
the shape of disaggregated container.
In addition to these different works, it can also be interesting to have a view of both
the external and internal structure of archaeological material in order to understand its
production process, or to spatially localize the different notable internal items. The goal
of the work presented in this paper is to propose a tangible representation of a complex
intricate structure through advanced 3D printing. The process is applied to a funeral urn,
and combines CT scan, 3D model processing, and 3D printing.
2 Archaeological Context
The excavation of the site of Domaine de la Bizaie in Guipry (Brittany, France, Fig. 1,
Left) in a preventive archaeological context uncovered a trapezoidal shaped funerary
enclosure. The central area housed ten cremation burials containing pottery vessels from
the Iron Age (Excavation L. Aubry, Inrap, [7]). The exceptional state of preservation of
some of these cremations prompted us to use tomography to analyse their contents. It
allowed to highlight a number of metal objects as in the F42A cremation (Fig. 1, Right).
Fig. 1. Left: Domaine de la Bizaie, Guipry, France. Right: the F42 cremation, in situ
434 T. Nicolas et al.
3 Description of the Work
The process used in this study is based on a combination of CT scan, 3D processing and
3D printing. The main goal of the process is to physically display the internal structure
of the archaeological material. We thus chose to explore two possibilities offered by 3D
printing: physical copy of a part of the initial material and global printing with trans‐
parency. The process follows these different steps:
1. A computed tomodensitometry of the archaeological material was performed gener‐
ating a database of X-Ray images, with density data.
2. Surface 3D meshes are generated from the previous database. Each mesh corre‐
sponds to a density range.
3. The 3D models are processed to fit 3D printing constraints.
4. The resulting 3D models are 3D printed.
3.1 Technical Environment
CT Scan. The CT scanner used in this study is a Siemens SOMATOM sensation 16

owned by the French company Image ET (http://www.image-et.fr/) who delivers serv‐
ices and expertise on X-Ray images and associated tools.
The three-dimensional explorations were performed through two passes (acquisi‐
tion) and 2D/3D images post-processing. The two acquisitions were a topogram (or radio
mode) for positioning the slices to be realized and a helix scan.
The scan generates a dataset under DICOM format [8] of 512 × 512 pixels 2D slice
images.
3D Data Processing. The data generated during the CT scan was processed with the
free software Osirix, an image processing application for Mac dedicated to DICOM
images. The 3D surface rendering functionality was used to generate 3D meshes corre‐
sponding to different density ranges. This functionality allows to generate the surface
of a set of points whose density value is within a defined range.
The 3D model was then manually processed in order to remove unwanted data and
automatically processed in order to get a ready-to print file.
3D Printing. We performed two different tasks of 3D printing on the urn, using two
technologies: 3D printing of an internal artefact, and 3D printing of a whole material
with transparency. The first 3D printer was a Stratasys Mojo owned by our institute.
This 3D printer uses Plastic Jet Printing to print objects from ABS, in layers as thin as
178μ, with a maximum dimension of 12.7 cm × 12.7 cm × 12.7 cm. For the second
printing, transparent 3D printing for complex objects appeared to be a not very developed
technique. We contacted several Companies delivering advanced 3D printing services
and only one, the CADindus Company (www.cadindus.fr), positively answered to our
request. It uses a technology of multimaterial and multicolor additive manufacturing by
resin polymerization for the production of the copies, on a Stratasys Objet500 Connex3
3D printer. This printer has a printing capacity of 49 cm × 35 cm × 25 cm, with an
accuracy of 30μ.
3.2 Application of the Process to the Urn

Scan of the Urn. The urn was scanned as a whole at one time. As the sediments inside
the urn were not too dense, the scan was performed with the values of 120 kV and
350 mAs. In order to scan the entire urn, the field of view was sized to 320 mm × 320 mm,
resulting in a resolution of 625μ. We worked in an extended Hounsfield scale (from
−10.000 to +40.000), in order to distinguish between fragments of bones and metallic
parts and get a finer view of the metallic objects. A volume rendering is presented in
Fig. 2.
Fig. 2. Volume rendering of the urn (Color figure online)
The three views of volume rendering of the urn presented in Fig. 2 highlight the
metal parts in blue, with two notable objects, a fibula and a knife blade. The red parts
are fragments of bones. The sediments are displayed in transparent grey.
3D Data Generation and Processing. Three meshes were generated from the data with
the Osirix software (Fig. 3). The first mesh, for the metallic parts, corresponds to the
points whose radio-density is between 4500 and 10950, the second mesh, for the urn
shape and sediments, between 700 and 1300, and the last mesh, for the bones parts,
between 1600 and 2300.
Fig. 3. Left: meshes for the urn and sediments, Middle: metallic parts, Right: bones
Two issues arose after the generation of the meshes. First, corrosion gangues on the
knife blade and fibula (Fig. 4, left) had a radio-density within the range of bones, visible
in the mesh of the bones (Fig. 3, right). These gangues were easily removed as they were
completely separated from other bones. Second, the mesh of the urn and ashes contained
many galleries and cavities due to worms and insects (Fig. 4, right). All the galleries
had to be manually removed and closed in order to get a clean transparency inside the
urn. This task took several weeks, with many exchanges with the archaeologists in order
to validate the modifications of the model.
Fig. 4. Left: gangue of corrosion for the metallic parts. Right: worms and insect galleries inside
the urn
The targeted impression technique for the full urn required filling the volume of the
urn, with a negative print of the metallic objects and bones fragments. This task was
performed with 3D Boolean operations in a 3D modeling tool.
Printing of the Fibula. The mesh of the fibula was isolated in order to propose a phys‐
ical virtual extraction of this object whose shape appeared to be very well preserved.
The process of producing the physical copy of the fibula from the digital data appeared
to be very simple and fast (less than one day). The resulting copy of the fibula is presented
in Fig. 5.
Fig. 5. 3D printing of the fibula
Printing of the Urn. In order to keep close to the initial volume renderings from
tomography (Fig. 2), we chose to render the bones in magenta and the metallic parts in
cyan. The printing phase took 50 h and required 8 kg of matter. It was printed with
successive layers, bottom up (Fig. 6, left). The resulting printed object was covered with
a pink support matter (Fig. 6, right)
Fig. 6. Left: on-going 3D printing of the urn. Right: the printed urn with its support matter (Color
figure online)
Post-processing. An important post-processing of the object produced by the 3D

printer is required to obtain a good transparency rendering (Fig. 7). This post processing,
which lasted one week, consisted in four phases: (i) Removal of the support matter, (ii)
Photo-bleaching during 72 h at 6500 kelvins («day light»), (iii) Sanding cycles with
different size grit (from 180 to 800), (iv) Coating in sterile conditions.
Fig. 7. Transparency view in the 3D printed urn
4 Discussions on the Results
Before its actual excavation, the cremation 42A was the subject of a CT and a 3D recon‐
struction. 3D volume rendering informs on the preservation of the burial, the urn and
the cremated bone block (its fragmentation and the deformations of the internal masses
bioturbations…). It allows the identification, location and orientation in space of each
artifact in a sustainable manner, the visualization and localization of the bone mass in
its entirety and the observation of the nature of the sediment’s components in the urn.
These elements offer the possibility of taking immediate precautionary measures before
any manual intervention. Furthermore, the digitization permits operations of “virtual”
manipulation of artifacts to increase the observation providing the first interpretations
before excavation. It also serves as supporting tangible “evidence” for elements that are
difficult to characterize in 2D. With the segmentation tools, it is possible to obtain an
independent 3D model of an artifact that can be virtually manipulated before its exca‐
vation. The additive techniques allow tangible handling and initial observations before
the provision of the original. This process allows the provision of information within
hours, which is not the case for the “operational chain” commonly implemented that can
take months if protective measures are implemented, prior to the study of the archaeo‐
logical material, as presented in Fig. 8.
Fig. 8. Timelines of digital (blue) and physical (red) excavations (Color figure online)
Fig. 9. Left: the 3D printed fibula Right: the real fibula, after its excavation, and before the
restoration process.
The CT scan provided information on the structural state of the object, and allowed
the identification of the cluster type of the fibula. Nevertheless, the observation of 2D
sections and 3D reconstruction did not make possible to truly characterize the number
of turns of the spring and the pin holder. Different segmentations were performed
resulting in a restitution of the object with an approximate resolution of the details, which
requires a systematic return to the real object in fine. This model has been the subject
of a 3D impression (Fig. 9, left) that allowed to corroborate a number of observations
from CT scans but also to confirm with certainty the number of turns of the spring and
the shape of the. Also of note is the tangible nature of 3D printing with respect to the
digital model. These comments have all been confirmed after the implementation of
protective measures (Figs. 9, right and 10, left). The characterization of the initial surface
condition during manual cleaning by a conservator revealed a setting of longitudinal

striations on the arc (Fig. 10, right). Only this manual intervention may allow detection
of such a setting very difficult to determine virtually on a small object.
Fig. 10. Left: the fibula after restoration. Right: details of the etched patterns on the fibula
The CT scan reveals the different elements using an arbitrary color application (bone
remains in red, metal in blue, biological indices in grey, etc.). For the neophyte, this
palette can disturb, but is quickly understood. For the anthropologist, while ordinarily
the information is supplied layer by layer in stratigraphic excavation, this comprehensive
3D visualization is more interactive. Translating from 2D to 3D, is none other than the
transition from virtual to real and requires each expert to integrate this type of imaging.
The complex 3D printing digital model overcomes this difficulty. After treatment, the
printing transparency provides a 3D print of the container and contents (here funeral-
deposit burned bones and iron objects knife rivet and fibula, Fig. 11).
Fig. 11. Front and bottom views of the 3D printed urn (Color figure online)
Printing on both its tangible nature allows “direct” access to information and the
physical handling of cremation deposit. It allows viewing and manipulation of the vessel
and its contents in full and with integrity. Printing allows visualization of burned bones
that are imperceptible when excavating (too small, fragile, or appearing negative, to be
taken), but also some well-preserved bones (at least here, a long bone) and the deter‐
mination of iron artefacts (here released from their mineralization). 3D printing also
provides information for the reconstruction of the funerary urn. This tangible medium
allows a manipulation/simple visualization to work on the analysis (such as the distri‐
bution of artifacts and the burial gestures) but also as a support for the excavation [2].
Indeed, 3D printing is the only tangible medium of context preserved after the excavation
of the incineration.
5 Conclusion and Future Works
For the archaeologist, the first asset of the scanner remains the immediacy of informa‐
tion, essential in the case of diagnostics where time is short. Well before the excavation
of a container (funeral vase…), imaging allows the establishment of a protocol, the
scheduling of specialists’ intervention, the possible solicitation of a conservator curator
to keep the fluidity of the scientific treatment of the remains [9].
In this process, the substitution of the original by certified copies can be attractive
for a valuation framework, given the fragility of unstable materials to exhibit. The
archaeological sensitive materials suffer during traveling. Accidents due to manipulation
(packaging, transport and installations) are inevitable. 3D printing offers promising
perspectives regarding the diversity of materials faithful to the look, the weight, the
texture of the original… From the image to the copy of an original, it takes an average
a half-day with an easy correction of the design since we can re-intervene in the template
file at any time. We also underline its importance in an educational setting with teaching
kits embellished with 3D restitutions subject to multiple manipulations, but also for
people with disabilities.
The processes used open up new exploratory research. For example, the possibility
of obtaining a model of old or recent bioturbations (earthworms galleries…) in a funerary
urn provides information on the presence of a perishable container and 3D printing of
a model provides access to “all” of the skeletal remains.
Acknowledgement. This project was funded by the CNRS Imag’In IRMA project, France.
References
1. Laycock, S.D., Bell, G.D., Corps, N., Mortimore, D.B., Cox, G., May, S., Finkel, I.: Using a
combination of micro-computed tomography, CAD and 3D printing techniques to reconstruct
incomplete 19th-century cantonese chess pieces. J. Comput. Cult. Herit. 7(4), Article 25,
February 2015
2. Haßmann, H., Heintges, T., Rasink, B., Winghart, S., Wulf, F.-W.: Der bronzezeitliche
Hortfund von Gessel, Stadt Syke, Landkreis Diepholz, Berichte zur Denkmalpflege in
Niedersachsen Veröffentlichung des Niedersächsischen Landesamtes für Denkmalpflege
Hameln: Niemeyer, vol. 32, no. 1, pp. 23–28 (2012)
3. Nicolas, T., Gaugne, R., Tavernier, C., Gouranton, V., Arnaldi, B.: Preservative approach to
study encased archaeological artefacts. In: Ioannides, M., Magnenat-Thalmann, N., Fink, E.,
Žarnić, R., Yen, A.-Y., Quak, E. (eds.) EuroMed 2014. LNCS, vol. 8740, pp. 332–341.
4. Nicolas, T., Gaugne, R., Tavernier, C., Petit, Q., Gouranton, V., Arnaldi, B.: Touching and
interacting with inaccessible cultural heritage. Presence: Teleoperators Virtual Environ. 24(3),
265–277 (2015). MIT Press
5. McKnight, L.M., Adams, J.E., Chamberlain, A., Atherton-Woolham, S.D., Bibb, R.J.:
Application of clinical imaging and 3D printing to the identification of anomalies in an ancient
Egyptian animal mummy. J. Archaeol. Sci. Rep. 3, 328–332 (2015)
6. Nicolas, T., Gaugne, R., Tavernier, C., Gouranton, V., Arnaldi, B.: Tomography and 3D
printing for the study of archaeological artefacts. In: Proceedings of the Conference New
Technologies for Cultural Heritage, Paris, musée du quai Branly, November 2014, Cambridge
Scholars Publishing (to appear)
7. Aubry, L., Le Puil-Texier, M.: Un enclos funéraire de l’âge du Fer, Domaine de La Bizaie,
Guipry, (Ille-et-Vilaine): rapport de fouilles, Cesson-Sévigné: Inrap GO, pp. 113–119 (2014)
8. Mustra, M., Delac, K., Grgic, M.: Overview of the DICOM standard. In: 50th International
Symposium ELMAR, pp. 39–44 (2008)
9. Biron, M., Hurtin, S., Nicolas, T., Tavernier, C.: La tomographie des objets archéologiques
complexes et/ou altérés: outil d’identification, d’analyse et d’aide à la décision pour les
mesures conservatoires. In: Restaurer l’ordinaire. Exposer l’extraordinaire, Journée des
Restaurateurs en Archéologie (JRA), Arles (2014)
10. Mitsopoulou, V., Michailidis, D., Theodorou, E., Isidorou, S., Roussiakis, S., Vasilopoulos,
T., Polydoras, S., Kaisarlis, G., Spitas, V., Stathopoulou, E., Provatidis, C., Theodorou, G.:
Digitizing, modelling and 3D printing of skeletal digital models of Palaeoloxodon tiliensis
(Tilos, Dodecanese, Greece). Quat. Int. 379, 4–13 (2015)
Towards Monuments’ Holistic Digital
Documentation: the Saint Neophytos
Enkleistriotis Case Study
Marinos Ioannides1, Charalabos Ioannidis2,

Archimadrite Neophytos Enkleistriotis3, David Castrillo1,
Pavlos Chatzigrigoriou1(&), Eirini Papageorgiou1,
Georgios Leventis1, Vasiliki Nikolakopoulou1, Vasilis Athanasiou1,
Fotis Bourexis2, Sofia Soile2, Styliani Verykokou2,
Maria Costi de Castrillo1, and Christian Sovis4
1
Cyprus University of Technology,
31 Arch. Kyprianou Str., 3036 Limassol, Cyprus
{marinos.ioannides,p.chatzigrigoriou,e.papageorgiou,
georgios.leventis,v.nikolakopoulou,
vasilis.athanasiou}@cut.ac.cy,
[email protected],
[email protected]
2
National Technical University of Athens,
9 Iroon Polytechniou Str., 15780 Athens, Greece
{cioannid,ssoile}@survey.ntua.gr,
3
The Holy, Royal and Stavropegic Monastery of Saint Neophytos,
Tala, 60289 Paphos, Cyprus
[email protected]
4
7Reasons, Hauptplatz 11, 3462 Absdorf, Austria
[email protected]
Abstract. The expansion of the term “monument” to include the surrounding

area of the tangible cultural asset, its natural environment as well as the intan-
gible data relating to its existence and use has gradually resulted the formation
of the term “cultural landscapes”. “Monument” has evolved into “monumental
place” and a “place with its own soul” and nowadays into a “unity” incorpo-
rating the multiple and diversified views which regard the one and single object,
the cultural asset. In this paper and through the presented case study of Saint
Neophytos Enkleistriotis monument, we attempt to move further on, from the
view of the “unity” and the interdiscipilinary approach to the “holistic” view,
treating the cultural asset as a “whole”; a “whole” which will have been created
from the harmonious merge of all the multifaced entities of which it is
comprised.
Keywords: Holistic documentation 3D model Geometric documentation

Orthoimage Structure from motion UAV

DOI: 10.1007/978-3-319-48496-9_36
Towards Monuments’ Holistic Digital Documentation 443
1 Introduction
The emergence of a holistic approach to monument documentation is inextricably

associated with the term ‘monument’, the quest of its meaning and its interpretation to
the people involved in the documentation of the past. The term ‘monument’ has been
transmitted many times following the evolutionary expansion of the concept of ‘her-
itage’ and its semantical transfer into ‘Cultural Heritage’ (CH). The main reason was
the necessity to set tangible heritage in its wider context, and relate it with its physical
environment and its intangible aspects (spiritual, political and social values) [1–3].
Both the existing natural and cultural diversity among the nations of the world and the
appreciation of intangible heritage -as the primer fundamental element before its
materialization into tangible heritage- led to a more anthropological, holistic heritage
approach. Moreover, the infiltration of culture and nature in the concept of ‘cultural
landscape’ as well as the broadened shift towards viewing historic buildings as an
inseparable part of historic cities and sites, revealed the matter of ‘monumental place’.
In this way, a monument can be seen as a whole, within a place with a soul – a genius
loci [4].
Monumental places are places of cultural identity and of great human achievement.
While the tangible part of a monument may leave its material evidence in the course of
time, the intangible part, which is more closely related to its creator, may never survive.
The survival of intangible heritage is then depended on the onward (mainly oral)
transmission of the knowledge behind its creation and the values it reflects: historical
values, symbolical values, authenticity and quality. It is a transmission of practices,
knowledge and skills that encapsulate a form of ‘memory’ of the built heritage and its
relationship between the society and a normative system of ideas and beliefs. Conse-
quently, ‘values’ and ‘memory’ are the substances that transform a built structure into a
monument and are only a part of what the holistic documentation approach aims to
capture. Due to the variety of its individual components (both tangible and intangible,
material and immaterial, concrete and abstract, natural or artificial), monument docu-
mentation demands the coincidence of the effort of many scientific, engineering, and
management disciplines to become a reality in order to gather all these kinds of
professional data. Hence, interdisciplinary research and cross-sectorial approaches
among the CH domain, that integrate knowledge and methods using real synthesis, can
lead to the augmentation of documentation tools and standards in establishing holistic
practices. Data collection for monumental buildings, complexes, settlements, areas,
artifacts during documentation deals with “visible” data. This includes the documen-
tation of past indications about monument’s remains through geospatial surveying of its
form, the recording of the materials used for its first creation and possible later inter-
ventions as well as the construction technology, which can lead to adequate pathology
analysis and effective conservation of the building [5]. Along with the visible data, the
“invisible” data, such as occurred past events, stories of the lives of the users and the
local community, songs and lyrics written about it, liturgies in the case of religious
monuments etc., will constitute the finishing layers that would deliver a unified
information and knowledge hub.
444 M. Ioannides et al.
Extending narrow and sterile considerations, the approach does not seek to “freeze”
the intangible part of the monument, by reducing it into files, be in archives, inven-
tories, museums, libraries, text, audio and video records [2]. Besides, it aims to cre-
atively “revive” it by further exploiting every technological affordance of the present
era. Eventually, holistic approach for monuments’ digital documentation is oriented to
provide the framework -a basis- for the ongoing adjustment and evolvement of the
effective, multileveled and multi-perspective reuse of the documented information by
the diverse user groups involved in the creation, use and dissemination of digital
cultural heritage.
This paper focuses on holistic approach documentation activities partaken at St.
Neophytos church; a church with tremendous monumental interest both of its tangible
and intangible aspects as well as the unique blending of natural and man-made envi-
ronment that bears the signs of human work and the effect of multicriterial factors that
have shaped so far the history of the religious monument.
2 Methodology
The Digital Documentation of Monuments has met vast progress in recent years, since
it develops parallel to the Information, Communication and Technology (ICT) sector.
However, because of this fast development, digital documentation has also left many
gaps and created challenges on data quality, accuracy, standards, archiving, copyrights
and sustainability. There is a need to establish a methodology and a workflow, where
various users (experts and non-experts) should follow, aiming to maximize collabo-
ration, distribution and -at the end- capitalize the digital information of the monument.
The proposed methodology of holistic approach presents a workflow that could (at the
full end level) answer the requirements of the monuments’ digital documentation,
filling the above mentioned gaps, by involving all user needs from the beginning of the
process. Counting users’ needs on planning and implementing the holistic digital
documentation is crucial due to the fact that the documentation sector as well as the
whole life cycle of a digital cultural heritage asset, are by nature multidisciplinary (and
during the documentation process become interdisciplinary). Therefore, before decid-
ing “what kind of data do I need (from them)” it is essential to have an integrated
image of the potential users’ needs; their interrelationships, their intersectorial agendas
(in which phase of the documentation process do they meet), the problems they are
coming up against during the process and how they overcome those. This feedback,
together with the required types of data, will enrich the metadata (data about data) and
organize the semantics (information that needs to be interlinked) in order to shape
system’s ontology (monument’s conceptual model), countering the above mentioned
challenges (data quality, accuracy, standards, intellectual properties and sustainability).
The first step towards the proposed methodology of holistic digital documentation
is the analysis of cultural heritage users and stakeholders. They can be categorized in
two major teams: experts and non-experts [6] (Table 1); “experts” are users that create
digital data from documenting monuments and “non-experts” are users that only use
these data. Nevertheless, due to the fast development of the creative sector -especially
in EU- the “non-experts” are creating new content that we cannot afford to leave
outside the monuments’ digital documentation, especially in the holistic approach.

Hence, the boundaries of these teams are loose, leaving many non-experts to become
experts due to their engagement, knowledge and creativity on specific cultural heritage
assets (Fig. 1). On our proposed approach, expanding the boundaries between these
teams is our aim; including more non-expert users in planning and digital documenting
monuments augments the results, increases the coverage of the intangible heritage
aspect, minimizes the possibility of faults and maximizes the prospects of reusing the
digital data.
Table 1. Users’ teams (expert and non-experts)

Experts Non-experts
Anthropologists/Sociologists Economists
Archaeologists Electrical, Acoustic, Thermal Engineers
Architects Exhibition designers
Archivists/Librarians Fellows
Biologists General Public
Civil Engineers Tourists
Chemical Engineers Students
Computer Scientists/ICT Specialists Public Relation specialists
Conservators Advertisers
Curators Religion Authorities
Geologists Administrative & Territorial Institutes
Geographers Guides
Historians Multimedia specialists
Material Engineers Game developers
Structural Engineers Decision makers
Geomatics Engineers Police & Fire authority
Musicians/Artists Owner
Neuroscientists (Psychologists) Travel agents
Site managers Local associations
Lawyers/Legal specialists Local and Regional authorities
Physics NGO’s
Surveyors Non-profit organisations
The second step is a parallel analysis of the cultural heritage information that we
need to extract from the monuments’ digital documentation. According to Core Data
Index for historic buildings and monuments of the architectural heritage and its basic
aim, a possible classification of individual buildings and sites is by: name, location,
functional type, date, architect, building materials and techniques, physical condition,
and protection status [7]. In addition, it should be taken under consideration the fact
that depending on the country or the organization, the need for deeper levels of
architectural, archaeological, environmental, historical, and planning information will
vary; each one must define its own specific requirements. Towards the digital holistic
Fig. 1. Experts and non-experts users’ engagement in holistic digital documentation
approach, the data (and metadata) that we need to document should expand and adjust
to the quality restrictions given by ICT experts. A holistic unity of intellectual
frameworks and approaches of information gathering beyond the disciplinary per-
spectives, is crucial for the sufficient implementation of an effective, transdisciplinary
and collaborative work methodology that aims to define data collection process, case
study setup, selection and utilization of systems and instruments, knowledge man-
agement, implementation of semantically enriched models, exploitation in education
and business and the continuous advancement of the information gathered and reused.
In Fig. 2, the proposed data and metadata scheme clearly presents the necessity for a
horizontal connection of Digital Documentation with all data (and metadata) categories
[5]. As digitalization is a procedure that could be applied in analogue data, it is
necessary to clarify that the proposed holistic methodology requires the plethora of the
documentation to be digital and not digitized. That said, obviously we can’t avoid
analogue data, especially during historic research, so digitalization will be applied
following the same scheme of metadata. Last but not least, the 3D documentation of the
monument is essential for holistic approach; documenting and creating the 3D model is
a procedure that involves a number of users (experts and non-experts) that can directly
benefit from the information covered by the 3D and many of them could (and should)
offer data to upgrade the quality and enrich the information of the 3D model.
Fig. 2. Basic data and parameters documentation for holistic digital approach
In the case study of Saint Neophytos, the proposed methodology workflow was
followed. At the beginning there was an effort to cover a number of users (experts and
non-experts) in the process of documentation (Fig. 1). Since the approach is digital,
various methods were discussed and the team developed a scheme of data collection
aiming towards the holistic digital documentation methodology. In the first phase of the
documentation the developing of the 3D Model was the crucial part of the workflow.
In Fig. 3 one can see the representation of how different users are involved to collect
the data from the monument to develop the 3D Model; at the same time, users absorb
the extracted information to create new content; this way many of them are upgraded
from non-expert users to expert users. An example of the high value of the 3D digital
model in our case study is the demand that the documented information should be
vastly useful for structural analysis (by Archaeologists, Historians, Architects & Civil
Engineers), for pathology and conservation state analysis (Civil Engineers, Conser-
vators and Architects), for hazard analysis (Geologists, Civil Engineers, Architects,
Policy Makers), for humanities analysis (Historians, Architects, Curators, Multimedia
developers) for intervention analysis (Architects, Civil Engineers, Historians,
Archaeologists, Electrical & Mechanical Engineers, Curators) and finally for monu-
ments value analysis (Historians, Architects, Archaeologists, Civil Engineers, Policy
Makers, Curators, Stake Holders, Conservators).
Fig. 3. Different Users’ involvement in holistic digital documentation of the monument of Saint
Neophytos.
The outcome of the digital documentation of the Enkleistra of Saint Neophytos is a

highly-detailed 3D model, which may act as a reference point for the creation of
applications in the field of Information Technology (IT). The specific 3D model pro-
vides incentives to developers in creating applications that will exploit model’s full
potentials in order to become useful tools to either experts or non-experts. However, the
proper function of these applications relies on the existence of a corresponding data-
base on a server and/or on cloud systems, where the acquired data needs to be stored in
a format that will be accessible and easily interpreted by all. Towards the scope of
proper interpretation, it is essential to establish standards for the collected data, leading
this way to a qualitative data collection with improved integrity and decreased number
of stored redundant data. Taking into consideration database’s management, the stored
data should follow the “Open Access” principles in order for its wider access and reuse
to be ensured as different disciplines may enrich the database or even fulfill missing
data. Parallel to that, a user friendly interface should be developed to help non-expert
users to find the required information. The case study presented in this paper, is the first
approach to Digital Holistic Documentation of Monument, using the proposed work-
flow. Obviously the project is ongoing and we are expecting interesting results, eval-
uation and limitations after the full deployment of user’s engagement.
3 The Saint Neophytos Enkleistriotis Case Study
3.1 Values’ Documentation

The Enkleistra of Saint Neophytos is an outstanding testimony of the Byzantine art and
civilization at probably the most troubled and intense moments in this part of the
Mediterranean: crusaders (English, Normans), byzantine and Arab elements are clearly
reflected in the Enkleistra. The paintings (frescos) are characteristic of the “Komne-
nian” Byzantine art period (mid 12 century); one can identify strong emotions, mainly
of passion and suffering, sophisticated proportions and movements of the figures who
are dressed in rich clothes with wavy edges. There aren’t many frescos saved in such
good conservation state as the ones in the monument of Saint Neophytos Enkleistriotis.
The preservation of the Enkleistra and the extensive documentation of its origin,
development and further interventions provide clear information for its interpretation
and confirm the authenticity of each element. This existing written documentation, the
structure, organization and paintings of the Enkleistra are key to understanding the
ways of the monastic life in the twelfth century.
The organic form of the Enkleistra and the plasticity of the cave carved in the rock
differentiates it from any other religious building around the island. It is situated in a
unique landscape, excavated in a cliff overlooking a valley and has withstood land-
slides and earthquakes for more than 800 years. It has additionally been continuously
used for monastic life and as a pilgrimage destination for centuries. These factors add to
its historic and cultural value and attract people from all the corners of the globe until
this day. Function, symbolism, use and the natural shape of the mountain are inter-
connected in this monument, unified and understood into a unique example of Cypriot
religious architecture.
The history, the uniqueness of the structural form and the amazing frescoes, make
the Enkleistra of Saint Neophytos a monument of an exceptional tangible and intan-
gible cultural Cypriot -and world- heritage.
3.2 Geographical Documentation

The Enkleistra is situated in the southwest part of the island of Cyprus, north of Paphos,
6 km away from the sea, at an altitude of 450 m. above sea level; a strategic position
which offered control over the coast (Fig. 4). The monastery complex is located in a
Fig. 4. Geolocation of the Saint Neophytos’ Enkleistriotis monastery complex in Cyprus, north
of Paphos (left); the New Monastery (centre); the Old Monastery (right)
natural valley oriented south. There is a small waterfall on the northern side of the cliff
that continues into a stream running south. The Enkleistra is situated on its west
part. The mountain cliffs are of a white calcareous rock, a workable material which is
easy to be eroded by natural phenomena (rain, vegetation, landslides) and/or hewed by
man, thus facilitating the creation (natural or artificial) of caves.
3.3 Historical Documentation

Saint Neophytos Enkleistros (“The Recluse”) (1134 A.D.–1219 A.D.), was born in the
town of Lefkara, near Larnaca. His life runs parallel in time with one of the most
unstable political periods in Cyprus [8]. At the age of 17 years old he chose the
monastic life and for five years remained at the St. Chrysostom monastery near
Kyrenia, cultivating the land and learning how to read and write. After two years he set
off for Jerusalem and shortly afterwards attempted to go to Asia Minor in search of
solitary life and enlightment. His attempt was unsuccessful and therefore decided to
become an ascetic at the hilly area above Paphos [8, 9]. In 1159 A.D. he departed for
his place of reclusion and formed his cell by modifying an already existing one,
encurved on the body of the mountain.
In this section there will also be incorporated various information regarding the
stakeholders, testimonies, bibliography, possible past uses as well as legal status.
3.4 Architectural Documentation

Ever since 1159 A.D. Saint Neofytos spend the rest of his life in Enkleistra, thus he had
furtherly modified the place, gradually turning Enkleistra from a single cave into a
complex of caves and constructions (Figs. 5 and 6). Between 1159 and 1160 [8] Saint
Neophytos dug two caves: his Cell, in which he used to live until 1197 A.D. and where
his grave also lies, and a consecutive-to-the-cell cave with an altar (Bema), which was
dedicated to the Holy Cross. Both of these caves constitute the “First Enkleistra”.
Fig. 5. Sketch of the Saint Neophytos’ Enkleistriotis monastery complex in 1735 A.D.; the New
Monastery (centre); the Old Monastery (left) (sketch: adapted by Barskij)
Fig. 6. Plan of various caves in the greater area of the old monastery complex - Level 1 (center),
Level 3 (left) (plan: adapted by [8])
In 1170 A.D., due to his growing popularity, Saint Neophytos was obliged to
accept a disciple, establish a monastery and adjust the mountain structure/rock con-
structions to monastic rules. Consequently, the First Enkleistra got extended. This
extension included the excavation of a third cave towards northeast (Naos) around
1183 A.D. [8] and the construction of an extra space on the eastern side (Exonarthex),
in order for the functional causes of the orthodox church to be fulfilled. Probably for
functional reasons of the monastery the cave of the Refectory (dining room) was also
excavated on the north side of the three-cave unity.
Due to the increasing number of pilgrims Saint Neophytos dug another cave at a
higher level (the New Zion or Higher Enkleistra), which was completed by the end of
1197 A.D. and in which he moved and spent the rest of his life until his death around
1220 A.D. in search of solitude. Gradually more spaces have been added, thus
expanding the complex of the old troglodyte monastery. A closed room with a dome
was constructed above the Exonarthex, the Sacristy, another closed room at a higher
level of the Sacristy, the Holy Attendance and Sanctuary or Hagiasterion which is
connected to the Naos through a narrow hole on the floor, as well as one more cave on
the northern part of the cliff, above the Refectory (probably formed in 1170 A.D.), the
Hermitage or Saint John Baptist Cave, which is devoted to the Holy Cross.
The old troglodyte monastery is composed of five different levels (Fig. 7): Level 0
(level of the bridge passing over the west stream), Level 1 (approximately 5 m above
Level 0) where the Narthex, Naos, Bema, the Cell of the Saint and the Refectory are
situated, Level 2 (Sacristy), Level 3 (Hagiasterion), Level 4 (the New Zion cell and the
Saint John Baptist Cave). Because on the northeast (NE) side of the cliff there is a
waterfall and a stream, the old monastery has expanded in the late twelfth century
towards the southwest (SW), and as shown in the sketch made by Barskij (Fig. 5),
some constructions already existed in 1735 A.D., possibly destroyed by a subsequent
landslide. However, further to the SW, vestiges of other caves are traced, as well as
stone building constructions used until not long ago by shepherds.
Fig. 7. Facade of the old monastery complex with all five levels (photo: David Castrillo, 2016)
In this section there will also be incorporated various information regarding the
general typology, compared with other caves on a local and international level, possible
evolution of typology, detailed architectural drawings on various scales based on the
photogrammetric documentation.
3.5 Intervention Documentation

During the years 1963–1965 interventions of consolidation of the structure were carried
out with the supervision of the Department of Antiquities of Cyprus. A covered ver-
anda and five arches resembling the originals cited by the Saint were constructed (Level
1). Also stairs and paths from the Hagiasterion (Level 3) to the upper level (Level 4)
were constructed.
In this section more detailed intervention data will be gradually added, in addition
to a reference regarding intervention’s materials and interventions’ general assessment
(actually solved problems or created new ones, successful or not successful, etc.).
3.6 3D Model Documentation

An important stage is the detailed documentation through the 3D recording of the
monument; the products of this stage will be also used as the basemap for the studies of
other specialists. First, the complete 3D geometric documentation of the Enkleistra via
photogrammetric and surveying techniques is a fundamental step for addressing its
structural and environmental problems. The results of the geometric documentation
(sections, ortho-images, 3D models) will be provided to for its preservation and pro-
tection, as well as the planning of conservative repair operations. Furthermore, the
generation of 3D printings or photorealistic 3D tours depicting in detail the internal and
external parts of the Enkleistra is another objective of the 3D recording procedure and
promotion of the monument.
3.6.1 Field Survey – Data Acquisition

As far as the geometric documentation of the Enkleistra is concerned, a combination of
field surveying, terrestrial laser scanning, photogrammetric and computer vision
techniques was applied. First, a geodetic network was established connecting the area
in front of the rock, which includes the Enkleistra and the interior of all the rooms that
exist inside the rock; so, all measurements and data collected can have a common
georeference.
Scanning by a terrestrial time-of-flight laser scanner and image taking through
terrestrial photography and the use of a small quadcopter UAV have been made at the
external and surrounding area of the monument. A Canon EOS-1Ds Mark III camera
was used for the acquisition of the terrestrial images (5616 3744 each) using two
lenses of 24 mm and 16 mm focal length. Using the UAV, videos sequences depicting
the external façade of the rock were taken by a GoPro camera, from a distance of 5 to
10 m from the object.
At the 4 rooms of the lower part of the Enkleistra, i.e., the old Enkleistra, which
consists of the Naos dedicated to the Holy Cross (2 connected rooms) and the Cell of
Saint Neophytos with his tomb, as well as the fourth separated room next to the Saint’s
cell, and at the outdoor corridor from which doors lead into the lower part of the
Enkleistra, multiple laser scans and the acquisition of more than 1,500 terrestrial
images took place. The combined use of images, for the application of image-based 3D
modelling techniques, and laser scanning was necessary for the complete coverage of
the complex areas inside these rooms (e.g., inside the tomb of the Saint, under and
around the Altar in the Bema and the table in the cell, inside recesses in the walls of the
Enkleistra, etc.) In the rooms of the upper part of the Enkleistra (Level 4) only the
acquisition of photos (a few dozens for each room) was sufficient for the creation of 3D
models.
More than 200 ground control points (GCPs), including both artificial targets and
natural features easily recognizable in the imagery, were measured using Total Station
equipment for the georeferencing of the images; also, the retroreflective targets of the
laser scanner were measured to co-register the image-based and laser-based point
clouds.
3.6.2 Data Processing

The automatic creation of 3D models using image sequences relies on photogrammetric
and computer vision algorithms. The first step of the image-based modelling pipeline is
the Structure from Motion (SfM) process [10], which refers to the method of simul-
taneous computation of the camera six degrees-of-freedom poses (i.e., the camera
exterior orientation of the images) and the sparse 3D geometry of the scene. Feature
extraction algorithms, like SIFT [11] or SURF [12], image matching and robust outlier
rejection techniques [13] are used, usually in combination with a sequential (incre-
mental) algorithm [12, 13] for metric reconstruction, in case of calibrated cameras, or
projective reconstruction, in case of unknown interior orientation; in case of uncali-
brated cameras an auto-calibration process is implemented [14–16]. The georeferencing
of the SfM outputs is generally performed by estimating the 3D similarity transfor-
mation between the arbitrary SfM coordinate system and the coordinate reference
system using GCPs [17]. The process of dense image matching [18] is applied for the
creation of a dense 3D point cloud using the camera exterior and interior orientation of
the imagery, estimated by the SfM process. Then, a 3D mesh model can be created
using the generated dense point cloud [19] via a surface reconstruction algorithm.
Finally, texture mapping may be applied to the mesh model using the oriented images.
The result of the image-based 3D modelling process is a photorealistic textured 3D
surface model of the scene depicted in the imagery that refers to the coordinate system
defined by the GCPs or the arbitrary SfM coordinate system, in case of lack of such
kind of information.
The Agisoft PhotoScan software was used for the creation of dense point clouds
using images taken by the Canon EOS-1Ds Mark III camera as well as frames of the
videos taken by the GoPro camera. The processing of each internal room of the
Enkleistra was performed separately. Similarly, the outdoor corridor and the external
facade of the Enkleistra were divided into separate projects. The alignment of the
images was the first step performed by PhotoScan; it is a SfM process that generates a
sparse point cloud of the scene and computes the camera interior and exterior orien-
tation, using a modification of the SIFT algorithm for the extraction of feature points.
GCPs were manually measured in the corresponding images to transform the camera
poses and the sparse point clouds into the local coordinate reference system. Fur-
thermore, some regions of a few images (e.g., retroreflective targets, obstacles due to
measuring equipment) were masked in the images, in order to reduce the subsequent
editing of the resulting dense point clouds. Dense image matching was the next step
applied through PhotoScan; the software calculates depth information for each camera
and combines it into a single dense point cloud.
Due to the complexity of the Enkleistra, some small areas without cloud points
were observed in the generated 3D dense point clouds (i.e., under the Altar, inside
recesses, etc.). 3D information for such areas was derived by the point clouds generated
by the laser scanner. The processing of the scans was performed via the Leica Cyclone
software, including their registration via the common targets as well as their georef-
erencing, using the measured coordinates of the targets. The Geomagic Studio was
used for the registration and merging of the laser-based and image-based point clouds,
separately for each room of the Enkleistra, as well as their editing, including the
removal of wrong and noisy cloud points. Furthermore, this software was used for
surface reconstruction for each separate point cloud, resulting in 3D mesh surfaces for
each room of the Enkleistra. In areas where the image-based point cloud was sufficient
(i.e., upper part of the Enkleistra, corridor and exterior regions), only the editing of the
point cloud and the meshing were performed via Geomagic software.
The reconstructed surfaces were inserted again into the PhotoScan environment for
texture mapping, using the oriented images. The outputs of these steps are 3D textured
models for every internal and external part of the Enkleistra. Finally, the textured mesh
surfaces were merged via the Geomagic software. Except for 3D products, orthoimages
at a resolution of 1 mm were generated via PhotoScan software, using the oriented
images and the derived point clouds, by manually determining the level of section for
each orthoimage. The orthoimages as well as the texture atlases of the 3D models were
processed via the Adobe Photoshop software.
3.6.3 The 3D Model

The main product of the geometric documentation of the Enkleistra monument is a
detailed and accurate 3D model which includes both the external areas (the rock and
the newer constructions, like the corridor, etc.) and the internals of the rooms (of the
upper and the lower part of the Enkleistra) georeferenced in a unique coordinate
system, so that their relative horizontal and vertical position will be correct. In the
following figures typical 3D views, ortho-images and horizontal and vertical sections of
the monument are illustrated (Figs. 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21,
22, 23, 24 and 25).
3.7 Structural Analysis

From a structural point of view, three different construction elements can be detected:
the mountain rock, the stone masonry and the concrete. The mountain rock serves as
back wall (western side) as well as ceiling of the caves, the stone masonry includes the
walls, constructed to close the eastern façades of the excavated caves and which
include openings for access or ventilation (doors and windows), while out of concrete
are constructed all the contemporary interventions, namely the veranda (roof, arches,
beams, columns) and possibly, the vertical reinforcement walls of the façade.
Regarding the structural condition, the latter additions in the period 1963–1965
(concrete beams, stone retaining walls) seem not to have helped to the stability of the
Fig. 8. Left: Projection on a single horizontal plane of the rooms of the Enkleistra that are at
different levels (Level 1 up to Level 4) and the footprints of the vertical sections (AA and BB);
Right: A view towards the ceiling at a horizontal section of the lower part of the Enkleistra
Fig. 9. The vertical section A–A′′′ through the rock that includes the Enkleistra, with a metric
scale of the relative elevations at the left (±0.00 m. at the entrance of Naos)
monument. On the contrary, it can be an added problem, since a theory for the dete-
riorated state of the monument is that below exist hollow areas (probably previous
rooms or auxiliary spaces like stables, etc. described by the saint in his writings). These
Fig. 10. Left: The vertical section BB′; Right: The façade of the rock that includes the Enkleistra
Fig. 11. 3D view from the outside of the inside surfaces of the four rooms of the lower part of
the Enkleistra (Level 1), as extracted from the point clouds of the laser scans.
voids, with the help of water infiltrated, seismic movements and the extra weight of the
new interventions, could imply the displacement of the masonry stone walls and the
evident detachment of the Narthex and the wall between them and also from the rest of
the structure (Fig. 3). Moreover, the use of mortars rich in cement adds to the dete-
rioration of the building structure.
In this section, the investigation of structure (description of bearing structure,
construction techniques, distribution of building materials etc.) as well as mapping of
structural state (e.g. mapping of structural cracks, disposition of elements, etc.) will
gradually be included.
Fig. 12. 3D of the view from the outside of the inside of the Naos dedicated to the Holy Cross
Fig. 13. 3D view from the inside of the 3D textured model of the Naos
Fig. 14. Orthoimage of the western wall of the Naos
3.8 Building Materials’ Characterization

This section will include information about the type of materials (e.g. stone, mortar,
ceramic, metal, etc.), their physical properties, their composition (e.g. chemical, pet-
rographic, mineralogical), their mechanical properties as well as their processing (e.g.
cost and availability, production, etc.).
3.9 External Hazards

The Enkleistra, being an integral part of the natural environment of the area, presents a
variety of problems mainly caused by landslides and the characteristics, properties and
morphology of the area, weather conditions, lack of adequate maintenance or intensive
human contact. Landslides could potentially change the morphology of the mountain
side and the quality of the rock allows it to break down easily, a phenomenon which is
accelerated by earthquakes, rain, fauna and flora creating an additional safety concern
for visitors. Earthquakes of high intensity, from 5 to 7° on the Richter scale, are
frequent in the Paphos area. One must also note that fire is a great potential threat of the
monument due to its location. The shape of the valley where it is placed creates a ‘no
exit’ situation in the event of a fire originating from the lower part of the valley. Winds
from the south would aid in the acceleration of the dimensions of such a fire.
The large attraction of visitors results in the change of the interior atmospheric
conditions due to accumulation of people inside the monument as well as physical
contact and photographing of the frescoes. Additionally, interventions due to con-
temporary needs (electricity etc.) and maintenance are often damaging or irreversible
Fig. 15. 3D views from the outside of the inside of the Bema
Fig. 16. Orthoimages of Bema Left: the northwestern wall - Right: the eastern wall
Fig. 17. 3D view from the outside of the inside of the Cell of Saint Neophytos with his grave
due to use of inappropriate and incompatible materials and new elements (electrical
installation, lighting etc.).
In this section the environmental factors, the accidental actions, the impact of
particular human actions as well as the social and economic parameters will be further
on analyzed.
Fig. 18. Othoimages of the Cell of Saint Neophytos; Top: the eastern wall - Bottom: the western
wall
Fig. 19. 3D view of the textured model of the grave of Saint Neophytos
Fig. 20. 3D view from the outside of the inside of the Refectory
3.10 Conservation State – Pathology Analysis

The level of humidity in the cell is also quite high, thus numerous frescoes in the lower
parts of the walls have undergone extensive degradation. The main pathologies caused
by the aforementioned problems are missing or detaching frescoe parts evident in the
lower levels of the walls (frescoes degradation, large lacunas, lack of adherence and
hollow areas under frescoes) as well as growing cracks.
Fig. 21. Orthoimage of the inside of the Refectory of the Enkleistra
Fig. 22. 3D views of the interior of the Sacristy; Left: Eastern wall - Right: Northern wall
Cracks are mostly evident on masonry reinforced by new walls or concrete columns
(Fig. 26). The most recent earthquake in 2015 created displacements inside the cave
and further expanded the already existing cracks (Fig. 26). Given that the back wall
(mountain rock) does not show major damage and through the analysis of the cracks,
the theory developed is that these cracks were created due to cavities below the existing
cave. A more in-depth study reveals that the whole narthex is detached from the
masonry walls, and that the masonry walls are detached from the rock mountain. This
creates three separate structural elements which have different structural behavior.
Fig. 23. 3D views of the interior of the Hagiasterion of St. Neophytos; Left: Southern wall -
Right: Western wall
Fig. 24. 3D views of the interior of the New Zion; Left: Southern wall - Right: Western wall
More extensive research will reveal whether more caves (which seem to fail struc-
turally) exist below Level 0, causing this type of extensive cracking (Fig. 27).
The diagnosis of decay which is derived from the usage of visual observations,
non-destructive techniques, analytical testing, GIS decay mapping etc., the study of the
mechanism of decay and the diagnosis of vulnerability (e.g. static and dynamic
assessment, etc.) will be added to the provided aforementioned information.
3.11 Arts’ and Humanities’ Documentation

The architecture of the old monastery is inherently engaged with the layout of the
frescoes and the presented iconographical themes. The digitally documented tangible
structure is able to provide also a vast amount of intangible information, even life
Fig. 25. 3D views of the interior of the Hermitage or Saint John Baptist Cave; Left: southern
wall - Right: western wall future work
Fig. 26. Construction materials and structural conditions (architectural plan, David Castrillo
2016)
Fig. 27. Frescoes, conservation state with cracks and damaged areas (photos and graphic
recreation of frescoes: David Castrillo 2016).
experiences, by incorporating various philosophical, educational and pedagogical

meanings and interpretations.
In the case study of Saint Neophytos the 3D model provides detailed documen-
tation also of the frescoes that exist in the Cell, the Bema and the Naos. According to
written testimonies, the Enkleistra was painted in 1183 A.D. by Theodore Apseudes
while the final phase of the paintings, mainly in the Naos, belonged to a different phase,
after 1197 A.D. The paintings of the Enkleistra were completed and restored in 1503 A.
D. through the commission of another monk, called also Neophytos [8, 9, 20].
The iconographical themes that were selected to cover the surfaces of the afore-
mentioned spaces indicate that Enkleistra was meant to act as a “laboratory” for the
study of the human soul’s mechanisms and provide pilgrim with valuable pedagogical
lessons. The knowledge of these mechanism offers human the capability to intervene
upon them, in the pursue of victory upon the soul’s passions and weaknesses. The
pedagogical method of Saint Neophytos, which merges the content of the religious
scenes with the cave architecture, comprises of three levels that comply with the three
distinct zones of the iconographical programme.
The first level of the pedagogical training takes place inside the Cell of Saint
Neophytos. In this room, where he was spending his everyday life Saint Neophytos had
also engraved his own grave. The constant view of the grave trains the pilgrim about
the concept of “death” and the perception of death’s philosophical and spiritual
essence. It is not only about the concept of the “inevitable end” of all humans, but
furthermore about a training that sets human free, on one hand from the materialistic
assets, and on the other hand from his desire to prevail upon other humans. When the
pilgrim has reached to that point, then he experiences a grade of freedom that opens up
his soul and enables him to profoundly love God.
Inside the Bema takes place the second level of the pedagogical training, due to the
representation of Saint Neophytos between two Archangels, unique in the whole
byzantine iconography. This representation expresses the loving bond between the
human and his love object, namely God. But this road to Love and the unification of
human with God is full of traps, deception, plausible phenomena which hide the truth
and psychic projections of wishes (alike the projections of an oasis in the dessert),
which constitute the “enemies” of the soul. At the same time the way is presented by
which a human can cope with the aforementioned situations without damaging his own
soul.
The third and last level of the training takes place in Naos. In Naos there are three
zones of representations, the position and the general layout of the latter reveals to the
eyes of the pilgrim a full life-cycle, through the pedagogical aspect of which the
pilgrim can experience a profound spiritual situation.
The lowest zone of the iconographical programme presents ascetics as well as
kings, all of them with a halo around their heads, in order to teach that both those who
have abandoned the cosmic society and those who have reached the top of social
pyramid and governed the world can defeat the human passions and conquer holiness.
The presentation of the two opposites, the ones who have nothing (ascetics) and the
others who have everything (kings) conveys the message that it is possible for the rest
of the people who find themselves in the in-between zone to follow (if wanted) the path
of spiritual life and (still) succeed their goals.
The road to reach the ultimate goal (union with God) is represented in the middle
zone of the iconographical programme. It is the road of humility. This zone includes
scenes from the last days of Jesus Christ, from Holy Wednesday until the Holy Res-
urrection. It concerns the conscious decision of human to choose forbearance and
reconciliation with his co-humans through apologies.
The upper zone is covered by the scene of Ascension of Jesus Christ which is
placed on the “roof” of the Naos, extending around the narrow hole that connects Naos
with the Hagiasterion, from which Saint Neophytos used to provide the Holy Com-
munion to the pilgrims. In these scenes, the mortal human body of Jesus Christ is lifted
up to his Father, to be united with Him (God) and thus be transformed into immortal. In
this way Jesus Christ prevailed over death and general decay, which is the success of
the unification of humans with God.
Saint Neophytos used to provide the Holy Communion to the pilgrims through the
narrow hole on the floor of Hagiasterion. In order for the pilgrims to receive the Holy
Communion, they had to climb a ladder towards the narrow hole of Hagiasterion,
through which Saint Neophytos was expecting on his knees to provide it to them.
During this ascendance the pilgrims were experiencing initially, the approach to the
way of life of the people around Jesus Christ in the scene of Ascension, afterwards the
approach to Jesus Christ on the cloud presented in the upper zone of the iconographical
programme and finally the union with Him through the offer of his body and blood
(Holy Communion) executed by Saint Neophytos.
The aforementioned zones of iconographical programme with the architecture of
the cave in combination with the service held in the church and the “road” of ascedance
of each pilgrim towards the Holy Communion, thus the unification with God, were
providing an exceptional pedagogical lesson to the people, through the combined
cultivation of the senses, the feelings and the mind.
All the aforementioned aspects regard the intangible information that is incorporated in
the monument of Saint Neophytos, which can provide pedagogical content and experiences
to those who open up themselves to Christianity. They also serve as input for various
educational purposes. The challenge of reusing and creating new digitized cultural heritage
content is an issue that concerns scientists, practitioners, professionals and institutions in
the field of culture and education. On the one hand, arts and cultural heritage education has
been regarded as the key factors in development of the “knowledge society and creative
ability” [21]. On the other hand, the ongoing development of ICTs is providing new
opportunities of representation and interpretation of cultural heritage. Thus, learning about
cultural heritage with the mediation of ICT tools and methods, such as an e-learning
platform enriched with multimedia content, can be affective in all educational levels by
using appropriately the available ICT affordances [22].
The aim is to engage learners from all ages, underage learners, adolescents as well
as vocational trainers. Learning about the culture, the living conditions, the human
creations and attitudes in the course of time, understanding the historic-cultural events
and the evolution of society, and finally discerning what has survived at the socio-
cultural surroundings, is the quintessence of arts and cultural heritage education. But in
order to achieve the following learning approaches that take under consideration the
benefits of using ICTs in cultural heritage education, we need to have a clear image of
the range of the “knowledge” by which the present monument is dominated, so we can
adjust it to our wide range of target learners. Learning approaches such as: (1) per-
sonalized, inquiry-based learning (2) on-site and anywhere learning experiences
(3) interdisciplinary learning approaches and (4) collaborative learning experiences,
demand prior presence and on-site experiential investigation in order to perceive the
inherent value of Saint Neophytos monument, which will become the focus of the (e-)
learning course.
An on-site analysis on Enkleistra’s structure and its surrounding place was sup-
plementarily conducted by research team’s multimedia expert, while the required data
were acquired through the use of state-of-the-art technological equipment. Moreover, a
360 camera facilitated the wide capture of Enkleistra’s both internal and external
landscape on video. The latter one can be used further on in the creation of VR video,
while various audio recording devices were put near the cavern in order to capture the
sounds caused by touristic engagement. Also, for the purposes of content enrichment
several pictures along with video were taken, using not only smartphones with
high-resolution cameras but also a UAV (drone) equipped with the latest GoPro 4
camera too. These acquired data will be represented in a digital installation of a
physical interactive book, in which the story of Saint Neophytos will unfold before the
eyes of readers encouraging them to engage with it using their hands, thus providing an
educational approach of the monument’s state. In order to accomplish this, the use of a
computer software called VVVV is needed, where advanced programming skills in C#
are required for its development.
Targeting at all audiences and lifelong learning experiences that can promote
sustainable educational practices, we contribute to the survival of the monument’s
‘memory’ and ‘values’. In addition, by re-using its digital representation of its tangible
and intangible aspects in such a context, we further provide its ongoing virtual and oral
transmission of its history to the upcoming generations.
3.12 Digital Documentation’s Parameters

As the holistic approach that it is proposed in this paper is digital, a member of the
planning and research team is an ICT specialist. Having in mind the 3D models that are
produced using survey means as well as their derived data, all the aforementioned set of
information will be stored in an SQL relational database taking advantage of DB’s
capability in supporting plethora of data types (e.g. integer, chars, floats, strings)
constituting it an ideal tool that will act as a reference point for the applications to run
effectively. Additionally, this kind of database will be created on Microsoft’s SQL
Server, as it runs on programming extensions called Transact-SQL (T-SQL) expanding
the common SQL standard to include more functions for the processing of string
variables and procedural programming. Moreover, considering the fact that there may
be compatibility issues for groups of experts who use open-source technologies, a
MySQL database will be further considered as the primary common ground to be used.
An important application that will exploit such a DB constitutes the development of an
online learning system depicting the cultural significance of Enkleistra using CMS (e.g.
WordPress, Joomla etc.) along with PHP code and various plugins enhancing plat-
form’s usability, where users will have the opportunity to be educated through
interactive courses that will present the archived data in a user-friendly way. In addition
to the proposed website, a mobile application will be considered for development in an
IDE like Android Studio taking advantage of JAVA’s capabilities towards the provision
of useful information to users regarding Enkleistra’s history and state through the use
of 3D models. Assuming the existence of huge amount of data that will be generated
over time, the use of data mining techniques to the aforementioned applications will be
crucial for their analysis in order to improve their functionality. Towards that end, the
collection of sufficient training data is most needed as Artificial Intelligence (AI) has
reached a point, where advanced machine learning methods can be implemented in data
mining techniques enriching their functionality. The creation of ML scripts that will
either be built-in at the code or loaded externally using command(s) will provide the
application the ability to “learn” by discovering various patterns that may exist in the
data, targeting the enhancement of user experience and the designation of proper
preservation methods for Enkleistra.
A Holistic Documentation Approach to any Cultural Heritage element/monument

requires an extensive and interdisciplinary research that considers its tangible and
intangible aspects, complimented with a strong reflection in a wide variety of subjects.
This is a pioneer approach in current documentation methodologies in the field.
‘Building’ plus ‘memory’, tangible plus intangible, is what transforms a simple
structure into a monument, and this is what a holistic digital documentation aims to
capture. The Saint Neophytos case study investigates the building’s in-temporality
which waits to be deconstructed and constructed again into a new form of communi-
cation and will form pilot mechanism for other monuments. The application of a
holistic approach can be the key to guarantee the uniqueness of the monument, and the
documentation at several levels is the key to protect the cultural values of the place.
The approach followed focused on the documentation and the designation of
appropriate preservation methods for the Enkleistra of St. Neophytos monastery, since
it is a monument of great cultural significance. The outcome should act as a reference
point for the application of restoration techniques that should be applied on Enkleistra’s
physical infrastructure in order to limit the risks that threaten its integrity. Nowadays it
is difficult to finance these methods due to the economic crisis that exists in the
Mediterranean countries, however an important measure that should be taken into
consideration is the listing of Enkleistra on UNESCO’s World Heritage Catalogue as a
cultural heritage site of outstanding universal value from historical point of view by
meeting certain criteria that should apply on its tangible form [23], leading this way in
the raise of both domestic and global awareness towards its protection and
conservation.
As technology has entered rapidly in peoples’ daily lives, there are various digital
online learning systems aiming to disseminate both tangible and intangible significance
of historical monuments that are mentioned on UNESCO’s World Heritage Catalogue.
An online educational tool, created by [22], provides a new way of reusing Cultural
Heritage data obtained from image-based and 3D modelling, terrestrial laser scanning
techniques, photogrammetry methods etc. which could also be applied in St. Neo-
phytos case. The proposed platform includes interactive lessons and activities using 3D
representations of the cavern, where internet users are taught monuments’ historic
background in various multimedia ways (by solving digital puzzles that reenact
hagiographies, watching dramatized videos that present the key elements of the
monument). At the end of each lesson, users are asked to answer few comprehensive
questions regarding the undertaken course awarding them with a ready-to-print state-
ment of accomplishment. These online interactive courses could assist users in
acquiring a better understanding of the existing risks that threaten Enkleistra’s integrity.
Furthermore, informative dissemination, such as publications, books, exhibitions, as
well as re-creation of case study’s digital content through 3d printing, screening,
modeling, etc., will be helpful in the direction of raising public awareness about the
monument and the necessity for its immediate conservation.
Although the expected increase of cultural tourism may constitute a beneficial side
of raising awareness, it stimulates issues regarding its preservation as there are plenty
of tourists that would like to interact with the already damaged hagiographies creating
this way more detriment and degeneration on their surface. Fortunately, the use of
Virtual, Augmented and Mixed Reality technologies have reached a point of evolution
where they can offer realistic and interactive experiences to humans. Aiming at the
decrease of touristic engagement inside the cave, an exploitation of these technologies
can be achieved. More specifically, visitors would be “transferred” into a virtual rep-
resentation of the cave where the Saint spent all of his life and could interact with the
digital frescoes in a way that is both engaging and educational through their partici-
pation in game(s), where they would be asked to fill the missing hagiographic segments
in order for St. Neophytos to “come alive” and recite his story. Towards the afore-
mentioned, an interactive analogue book can be located in the monastery’s entrance
where tourists may experience another form of educational environment grasping all
the multimedia data (images, texts, 3D, videos etc.) that thrive from the monument
while at the same time would be presented in a user friendly way.
Last but not least, creation of software and storage of the extracted cultural data
along with its semantic interrelations, can be used in collaboration with modern
advanced information systems like HBIM in depicting the optimal methods that can be
exploited towards Enkleistra’s preservation. This kind of management system can act
as a reference point for the holistic digital documentation of this monumental cave,
providing reverse engineering data to professionals from multidisciplinary areas that
work together in the protection and preservation using detailed 3D models. By this
way, academic research between various disciplinary sectors will be assisted too, and
links with other methodological approaches and practices for monuments of similar
characteristics or of common elements can be established.
Acknowledgements. 1. ITN-DCH (http://www.itn-dch.eu/) project has received funding from

the European Union’s Seventh Framework program for research, technological development and
demonstration under GA n° 608013.
2. INCEPTION (http://www.inception-project.eu/) has received funding from the EU’s H2020
Reflective framework programme for research and innovation under GA n° 665220.
3. Europeana Space (http://www.europeana-space.eu/) has received funding from the Euro-

pean Union’s ICT Policy Support Program as part of the Competitiveness and Innovation
Framework Program, under GA n° 621037.
4. 4DCH (http://www.4d-ch-world.eu/) project has received funding from the European
Union’s Seventh Framework program for research, technological development and demonstra-
tion under GA n° 324523.
5. Lo-Cloud (http://www.locloud.eu/) is co-funded under the CIP ICT-PSP program under GA
n° 325099.
6. ViMM (http://www.vi-mm.eu/) has received funding from the EU’s H2020 framework
programme for support and coordination actions under GA n° 727107.
References
1. Ahmad, Y.: The scope and definitions of heritage: from tangible to intangible. Int. J. Herit.
Stud. 12, 292–300 (2006)
2. Bouchenaki, M.: A major advance towards a holistic approach to heritage conservation: the
2003 Intangible Heritage Convention. Int. J. Intang. Herit. 2, 106–109 (2007)
3. Vecco, M.: A definition of cultural heritage: from the tangible to the intangible. J. Cult.
Herit. 11, 321–324 (2010)
4. Rodwell, D.: Sustainability and the holistic approach to the conservation of historic cities.
J. Archit. Conserv. 9, 58–73 (2003)
5. INCEPTION Project: D.1.3 - Project Report (2016)
6. ITN-DCH Project: D.1.1 - Project Report (2014)
7. Bold, J., Chatenet, M., Cortembos, T., Kovacec Naglic, K.: Guidance on inventory and
documentation of the cultural heritage, Strasbourg (2009)
8. Mango, C., Hawkins, E.J.W.: The hermitage of St. Neophytos and its wall paintings.
Dumbarton Oaks Papers 20, p. 119 (1966)
9. Tsiknopoullos, I.P.: The Encleistra and Saint Neophytos, Nicosia (1965)
10. Hartley, R., Zisserman, A.: Multiple View Geometry in Computer Vision, 2nd edn.
Cambridge University Press, Cambridge (2003)
11. Lowe, D.G.: Distinctive image features from scale-invariant keypoints. Int. J. Comput. Vis.
60, 91–110 (2004)
12. Bay, H., Tuytelaars, T., Gool, L.: SURF: Speeded Up Robust Features. In: Leonardis, A.,
Bischof, H., Pinz, A. (eds.) ECCV 2006. LNCS, vol. 3951, pp. 404–417. Springer,
Heidelberg (2006). doi:10.1007/11744023_32
13. Fischler, M.A., Bolles, R.C.: Random sample consensus: a paradigm for model fitting with
applications to image analysis and automated cartography. Commun. ACM 24, 381–395
(1981)
14. Snavely, N., Seitz, S.M., Szeliski, R.: Photo tourism: exploring photo collections in 3D.
ACM Trans. Graph. 25, 835–846 (2006)
15. Klopschitz, M., Irschara, A., Reitmayr, G., Schmalstieg, D.: Robust incremental structure
from motion. In: Third International Symposium on 3D Data Processing, Visualization and
Transmission (2010)
16. Hemayed, E.E.: A survey of camera self-calibration. In: Proceedings of the IEEE Conference
on Advanced Video and Signal Based Surveillance, AVSS 2003, pp. 351–357 (2003)
17. Nilosek, D., Walvoord, D.J., Salvaggio, C.: Assessing geoaccuracy of structure from motion
point clouds from long-range image collections. Opt. Eng. 53, 113112 (2014)
18. Remondino, F., Spera, M.G., Nocerino, E., Menna, F., Nex, F., Gonizzi-Barsanti, S.: Dense
image matching: comparisons and analyses. In: Arqueologica 2.0, et al. (eds.) Proceedings of
the DigitalHeritage 2013 - Federating the 19th International VSMM, 10th Eurograph-
ics GCH, and 2nd UNESCO Memory of the World Conferences, Plus Special Sessions from
CAA, pp. 47–54 (2013)
19. Remondino, F.: From point cloud to surface: the modeling and visualization problem. Int.
Arch. Photogramm. Remote Sens. Spat. Inf. Sci. XXXIV, 24–28 (2003)
20. Kakoulli, I., Fischer, C.: The techniques and materials of the wall paintings at the Enkleistra
of St. Neophytos (Phase II) (2012)
21. Ott, M., Pozzi, F.: ICT and cultural heritage education: which added value? In: Lytras, M.D.,
Carroll, J.M., Damiani, E., Tennyson, R.D. (eds.) WSKS 2008. LNCS (LNAI), vol. 5288,
22. Ioannides, M., Bokolas, V., Chatzigrigoriou, P., Nikolakopoulou, V., Athanasiou, V.:
Education use of 3D models and photogrammetry content; the Europeana space project for
Cypriot UNESCO monuments. In: Proceedings of Fourth International Conference on
Remote Sensing and Geoinformation of Environment, Cyprus, April 2016
23. UNESCO: Operational Guidelines for the Implementation of the World Heritage Conven-
tion. WHC.02/2 (2002)
Project Paper: Data, Metadata,
Semantics and Ontologies in Cultural
Heritage
First Experiences of Applying a Model
Classification for Digital 3D Reconstruction
in the Context of Humanities Research
Sander Münster1(&), Cindy Kröber1, Wolfgang Hegel1,

Mieke Pfarr-Harfst2, Nikolas Prechtel3, Rainer Uhlemann1,
and Frank Henze1
1
Media Center, Technische Universität Dresden, Dresden, Germany
[email protected]
2
Department of Architectural Design, Technische Universität Darmstadt,
Darmstadt, Germany
3
Department of Cartography,
Technische Universität Dresden, Dresden, Germany
Abstract. While technological backgrounds, project opportunities, and

methodological considerations for application are widely discussed, there is still
no comprehensive classification scheme for digital 3D reconstruction in
humanities research projects. Therefore we developed a prototype scheme in
2016. In this article we present the first results of applying this scheme and
classifying five projects. Within this application we tested for intercoder relia-
bility and for potential weaknesses of the scheme. While the reliability of the
proposed scheme is generally good for categories with discrete values, quali-
tative categories result in highly differing coding.
Keywords: Cultural heritage Information management Classification

models Digital reconstruction
1 Introduction
Not untypically for research activities which cross disciplinary frontiers, it can be hard
to formalize 3D reconstructions using standard indicators, such as a joint research
object or explanatory approach. In contrast, considering the use of a common research
method seems promising [1]. Since 3D reconstruction processes foster a nomothetic
and holistic representation of the past, are they a methodological step backwards
compared to modern problem-centered and constructionist approaches in the human-
ities [c.f. 2]? The humanities are facing various challenges in their use of digital
reconstruction methods, which originate from architectural design, engineering, and
geosciences, and are highly reliant on tools from computing. Examples include the
need for simplification through model building, multiple authorship, or visual research
strategies. These methods have been commonly used in both the academic and the
private sector. Currently, digital reconstructions are mainly carried out in one single use
context by interdisciplinary workgroups applying expert technologies.

DOI: 10.1007/978-3-319-48496-9_37
Considering this background, the existence of numerous standards, guidelines and

rules for dealing with historical content [3–7], which have only achieved limited
practical relevance [8], has proven to be a challenge. In contrast, the concept of
metadata in classification and descriptions of historical information has become widely
established. Although one of these schemas, CIDOC-CRM [9], has gained popularity
as a reference ontology (in terms of a generic concept of knowledge structure) in
archaeology and museology to describe reconstructed objects, standards for metadata
and its implementation as a whole remain highly heterogeneous [10, 11]. Current
approaches to sustainable documentation of the creative process of digital recon-
structions have not yet become sufficiently established in practice, despite diverse and
innovative concepts [12, 13]. In this context, our interest is less in building another
metadata scheme than in identifying general properties which characterize digital 3D
reconstruction projects. Classification schemes need to be made easily applicable and
their results comparable. Quantifiable categories would be helpful to achieve this aim,
which relies on two concurring interests. On the one hand, an applicable classification
scheme has to provide all variables of relevance to describe a certain object and a
comprehensive set of values. On the other hand, an optimal scheme must be generic
and slim to make it easily applicable—bearing in mind that schemes are often disre-
garded for the high workload involved—and consistent, that is, equipped with a full set
of parameters. Moreover, an ideal classification scheme should also allow for classi-
fication ex post by non-insiders.
According to Merriam-Webster’s dictionary, classification is a “systematic
arrangement in groups or categories according to established criteria”.1 Classification
aims at creating a taxonomy for a certain branch or aspect of information [14]. An
ontology combines a taxonomy with the underlying principles of classification. Thus,
an ontology sorts individual items into a specific scheme according to defined rules.
Regarding the purpose of classification, it delivers “particular frame[s] of reference
[which] help[s] us to make sense (both in passive and active meanings) of the world”
and is widely discussed from philosophical as well as historical and practical points of
view [e.g. 15, 16]. In the context of scholarly work, classification schemes are closely
related to specific disciplines or—in a broader sense—epistemic cultures and supply a
shared understanding of research objects, habits, methodologies, or publication bodies.
This is an important prerequisite for further research-related tasks, such as the definition
of best practice, implementation within academic curricula, shaping the scope of
conference programs, and investigating funding and management responsibilities or
cooperation networks [17].
In 2016 the authors proposed a preliminary classification scheme for digital
reconstruction in humanities research [18]. Its development was based on the inves-
tigation of 2,931 journal articles and conference papers concerning the 3D recon-
struction of historic objects, employing deductive and inductive methods of content
analysis [19] and bibliometrics [20]. In order to examine project practices and their
evolution, a qualitative content analysis of a further 26 international publications
concerning digital reconstruction of lost objects was carried out. While such
1
http://www.merriam-webster.com/dictionary/classification, accessed May 11, 2016.
First Experiences of Applying a Model Classification 479
publications provide records of past projects, follow-up research into four case studies
was conducted and examined by a mixed-method approach, including heuristic
frameworks and Grounded Theory [21]. While the development of the classification
scheme and its elements were widely discussed in the previously mentioned article
[18], the main research question here is whether the proposed scheme can be applied
for various types of digital 3D reconstruction projects and handled by various indi-
viduals. So this article focuses on the following:
1. The application of the developed categorization scheme by different users in 6
example projects.
2. Evaluating the usability, eligibility, practicability and potential shortcomings of the
scheme.
2 A Preliminary Classification Scheme for Digital

Reconstruction in the Humanities
To briefly introduce the classification scheme, a related classification matrix distin-

guishes between (1) the research context in terms of the subject and the historic object
and (2) the quality of digital 3D reconstruction (Fig. 1). The quality pertains to sources
of reconstruction, technologies, the characteristics of the resultant computer model,
documentation of reconstruction processes and—if created by an interdisciplinary team
—a record of the collaboration between stakeholders.
Fig. 1. General structure of a classification scheme for digital reconstruction
Taking into consideration that digital 3D reconstruction is highly heterogeneous in

terms of objects, purposes, and workflows, we suggest a low-level compromise. Our
proposed scheme contains some mandatory and widely quantified categories which are
easily defined. A basic set of three variables has been identified: Technology (tools and
geometry data type), project cooperation (project scale and date), and historical object
(location, time of origin, and type of object). Most of the other descriptions (e.g., model
qualities or documentation) are optional, not standardized, and qualitative only. Table 1
summarizes this classification scheme, the contained categories and variables.
Table 1. Overview of the preliminary classification scheme
Category Variable Characteristics, examples

Sources type of source original (material) evidence, acquired
data, images, texts, “logical source”
inherent knowledge primary source, secondary source
Model quality geometrical fidelity accuracy, detail, coherence
radiometric fidelity properties, properties depending on
accuracy, detail, coherence
temporal fidelity properties, properties depending on
accuracy, detail, coherence
Documentation result documentation reference ontology, application ontology
process documentation approach
Technology technological domain GIS, VR, CAD, BIM
model genetic workflow semi-automated genesis, procedural
generators, human-driven modeling
approach
tools software
geometry data type point cloud, wireframe model, polygon
model, voxel model, parametric model
(program)
Project cooperation project scale employees, budget
project dates start and end date
involved competencies humanities, design, technologies
Historic object type of object static artifact, dynamic artifact
general object properties tangibility, temporality
date of origin year, period
l o ca t i on p la ce
Research context research object source, historic object, system
research function documentation, data quality assessment,
visualization, process investigation,
conceptualization, contextualization,
numeric analysis, hypothetical simula-
tion
3 Application and Evaluation of the Scheme
This article evaluates the proposed classification scheme in the context of its appli-
cation to various projects. Five people were asked to classify five projects (Fig. 2)
according to the proposed classification scheme and to provide feedback on their
experiences with an application of this scheme. To ensure independence from specific
disciplinary backgrounds, we tested with individuals holding degrees in various dis-
ciplines, including history, architecture, geosciences, education, and informatics. For a
categorization study, it is important to estimate whether application of a scheme by
different individuals would lead to similar results. The effects and testing methods of
this intercoder reliability are widely discussed especially in the context of social sci-
ences [22]. Common practices to reduce the variance in classification entries include:
Project 1: Uch Enmek Project 2: Xi’an – 1st screenshot Project 2: Xi’an – 2nd screenshot
Project 3: Zwinger 3D Project 4: GEPAM Project 5: Freiberg Cathedral
Fig. 2. Examples of visualization outcomes
1. Providing well discernible and discrete sets of variants for all categories.
2. Providing anchor examples highlighting the application of the scheme.
3. Prior training of all coders on the application of the scheme [c.f. 23].
Aiming at a high level of awareness in our scenario, we supplied testers with both the
classification scheme (Table 1) and published application examples. To evaluate dif-
ferences between individuals in using the scheme, one of the projects was indepen-
dently categorized by two individuals who were involved in the project (Project 1). As
previously stated, the interest is in the applicability of the scheme, not only for project
insiders. This issue was tested by asking project outsiders to describe projects only
from published documents (Projects 2 and 5). The testers were directly involved in the
other projects.
3.1 Project 1: Uch Enmek Ethno-Nature Park

The Altay Mountains boast outstanding cultural heritage. Remains of the once flour-
ishing Scythian culture rank highest among the archaeological heritage found there,
with the most famous being located in the Karakol and Ursul Valley. This is reflected in
a conservation area called Uch Enmek Ethno-Nature Park. Archaeological expertise
and data on the project’s activities have been provided through a close collaboration
with archaeologists from Ghent University. The working group led by Jean Bourgeois
has been active in the Altay for some 15 years [24]. Courtesy of the GeoEye Foun-
dation, the Dresden group was assigned four overlapping IKONOS high resolution
satellite scenes of the Uch Enmek conservation site, forming an important data source.
3.2 Project 2: Xi’an – Emperor’s Power in the Afterlife

The exhibition “Xi’an – Emperor’s Power in the Afterlife”, presenting the research
results of 15 years of cooperation between Chinese and German institutions and sci-
entists, was the occasion for a computer reconstruction and simulation that was
embedded over three years in the ongoing scientific project. The 100 × 140 km area
around the old imperial city of Xi’an is regarded as a cultural region in the 21st century.
So far, in this region, two imperial tombs of the Qin Dynasty, 18 grave sites from the
Han Dynasty, and eleven imperial tombs of the Tang Dynasty have been discovered
and archaeologically investigated. Therefore, one primary objective of the project was
to relay the immense spatial dimensions of the area and the almost unimaginable
variety and number of archaeological sites. Simultaneously, the goal was to include
details of individual exhibits based on the most recent, as yet unpublished excavations,
surveys, findings, and research on the emperor and imperial tomb cities.
3.3 Project 3: Back to the Future - Zwinger 3D

The aim of the project is to virtually present the historical development of the Dresden
Zwinger, which has experienced some 13 major phases in its 300 years of existence.
Some of these were actually built, whilst others were merely ideas and ambitions of
influential individuals. Models were presented as an animated film, which summarizes
the project in an attractive way and provides a starting point for subsequent animations
and illustrations on the construction history of the Dresden Zwinger (Table 2).
3.4 Project 4: GEPAM: Memorial Landscapes - Dresden and Terezín

as Places to Remember the Shoah
The cities of Dresden and Terezín share the tragic history of the persecution of Jewish
people during World War II. Memorials in both cities commemorate the victims of the
Shoah. The purpose of 3D town models in the GEPAM project is to communicate the
variety of documents and information concerning the Shoah and allow the user to
evaluate events in terms of their location. Education concerning the Holocaust com-
plements the process of remembrance. The target structure of the final presentation
environment will allow access to documents and records related to the persecution of the
Jews within a web-based city model and serve as a virtual memorial. The GEPAM
project is financed by the EFRE scheme of the European Union supporting cross-border
cooperation between the Czech Republic and Germany (Tables 3, 4 and 5).
3.5 Project 5: Freiberg Cathedral – Creating an Educational App

The project was a student seminar during which an app had to be created within an
educational project. The outcomes of the individual efforts were bundled into a manual
for interdisciplinary student projects. Supported by the Saxon Centre for Teaching and
Learning in Higher Education, this seminar was a collaboration project between the
Table 2. Application of the proposed classification scheme: Project 1 – 1st and 2nd Coder
Category Project 1: Uch Enmek Ethno-Nature Park – Project 1: Uch Enmek Ethno-Nature Park
First Coder (Involved) – Second Coder (Involved)
Sources
Type of source acquired data (satellite images, digital elevation excavation documentation (here: site type and
models (DEM)), excavation plans, contemporary extent), photographs, satellite imagery, aerial
photographs, physical remains photographs, topographic reference, scientific
texts, field visits
Inherent primary source with probability of changes: primary sources: contemporary 21/2D land-
knowledge kuragan (burial mounds) remains scape model compiled from topographic
secondary sources: excavation plans, satellite references and remote sensing imagery,
images archaeological site documentation from recent
archaeological field work (text, photographs)
secondary sources: scientific literature on
Scythian burial mounds (for reconstruction)
Model quality
Geometrical level of detail (LOD) 2 models and generalized LOD1 of rural landscape, LOD2 of rural
fidelity land use buildings (prototypes representations accord-
ing to building taxonomy), LOD2 of archaeo-
logical objects (kuragans)
Radiometric generalized color scheme NPR model (not aiming at true color represen-
fidelity tation)
Temporal fidelity no temporal changes shows present state only (time reference by
imagery), multi-temporal extension possible
Documentation
Process doc. textual within thesis periodical protocols
Results doc. not indexed by metadata XML classification scheme of landscape
objects
Technology
Technological GIS-based modeling of landscape, VR modeling GIS-based modeling as a basis, VR-modeling
domain of obstacles, based on acq. data for 3D objects (buildings and archaeological
objects)
Model genetic semi-automated modeling of landscape, human- GIS landscape model as a synthesis of map
workflow driven modeling of obstacles digitization and interpretation of remote
sensing imagery, 3D object generation by 3D
modeling software
Tools ArcGIS, SketchUp, Maxon Cinema 4D ArcGIS, SketchUp, Maxon Cinema 4D
Geometry data shapefiles, polygon models shapefiles, polygon models
type
Project Cooperation
Project scale ~800 person-hours (modeling and data) ~1500 person-hours (modeling only)
Project dates 2007-2014 2011-2014
Involved compe- geoscientists, archaeologists, historians geoscientists, archaeologists
tencies
Historic Site
Original location Uch Enmek , Russia Uch Enmek Conservation Site (Siberia,
Russian Federation)
Time of origin ~ 600 A.D. present situation, archaeological objects: 5th–
7th century BC
General object (partially) no longer extant tangible object kuragans in different states of conservation,
properties most of them destroyed by illicit digging
Type of object landscape with obstacles 3D landscape model including schematized
3D archaeological sites
Research context
Research function visualization, conceptualization of land use (e.g., documentation of sites, public attention and
forest, settlements, or kuragans) awareness, education, workflow development
for automated transfer from GIS-based land-
scape models to 3D interactive, clickable
model objects (POIs) with context balloons,
NPR landscape visualization
Table 3. Application of the proposed classification scheme: Project 2, 1st and 2nd Coder
Category Project 2: Xi’an – Emperor’s Power in the After- Project 2: Xi’an – Emperor’s Power in
life – First Coder (Involved) the Afterlife – Second Coder (not
involved)
Sources
Type of source historical texts, texts, photographs, plans, results of photographs, excavation plans
excavation, sketches, comparable buildings or build
structures, findings, data from archaeological pro-
spection
Inherent primary sources: remains of 11 buildings, 18 lost primary source: remains, excavation data
knowledge buildings, data from archaeological prospections secondary sources: interpretation by
secondary sources: excavation plans, sketches, texts, archaeologists
comparable buildings or structures, photographs,
discussion with archaeologist
- Model quality
Geometrical LOD0: model of region of Xián not applicable
fidelity LOD1: model of the complete structure of the whole
mausoleum with 3 parts
LOD2: models of the 3 parts of the mausoleum
(north, south, hill)
LOD3: models of 29 buildings
LOD4: detailed model of the tomb inside the hill
Radiometric detailed textured models textured models
fidelity
Temporal fidelity time span of Tang Dynasty not applicable
- Documentation
Process documen- published: http://tuprints.ulb.tu-darmstadt.de/2302/ published dissertation
tation
Results documen- published: http://tuprints.ulb.tu-darmstadt.de/2302/; exhibition
tation film projection
Technology
Technological not named VR modeling of buildings based on
domain sources
Model genetic human-driven modeling human-driven modeling of buildings
workflow
Tools Maya, After Effects, Adobe Premiere Maya
Geometry data polygon models polygon models
type
Project Cooperation
Project scale ~ 4,160 person-hours (with students) not applicable
Project dates 2004-2006 2005-2006
Involved compe- architects, archaeologists, geoscientists, historians architects, archaeologists
tencies
Historic Site
Original location Zhaoling, China Zhaoling, China
Time of origin ~ 6th-7th century AD ~ 6th-7th century AD
General object (Partially) no longer extant tangible object no longer extant tangible object
properties
Type of object landscape, city structures, buildings buildings
Research context
Research function digital reconstruction and simulation not applicable
Table 4. Application of the proposed classification scheme: Projects 3 and 4
Category Project 3: Back to the Future - Zwinger 3D Project 4: GEPAM Memorial Land-
scapes
Sources
Type of source text, pictures, photographs, plans, remaining build- text, pictures, photographs, plans, remain-
ings ing buildings
Inherent primary source: historical drawings, plans, sketches primary sources: 34 remaining buildings
knowledge and paintings with modifications, 58 lost buildings
secondary source: remaining buildings secondary historical sources: artistic
pictures, photographs, plans
secondary contemporary sources: photo-
graphs, plans, texts
Model quality
Geometrical LOD3 of buildings relevant for the subject, LOD1 of LOD1 of city landscape, LOD3 of build-
fidelity city buildings ings relevant for the subject
Radiometric architectural white paper color scheme generalized color scheme
fidelity
Temporal fidelity no temporal changes transitions: per year, binary (display/no
display)
Documentation
Process documen- no documentation periodical protocols and changelogs
tation
Results documen- no documentation XML metadata: time, author, location
tation for LOD 1 objects according to GIS
Technology
Technological human-driven modeling of LOD3 objects GIS-based modeling for LOD1 objects
domain GIS-based modeling of LOD1 objects VR-modeling for LOD 3 objects
Model genetic human-driven modeling of landscape human-driven modeling of obstacles
workflow
Tools Autodesk 3ds Max, Blender, Maxon Cinema 4D ArcGIS, SketchUp, Maxon Cinema 4D
Geometry data polygon models shapefiles, polygon models
type
Project Cooperation
Project scale ~7200 person-hours (modeling only) ~3200 person-hours (modeling only)
Project dates 2006-2012 2012-2014
Involved compe- historians, architects, information scientists geoscientists, information scientists,
tencies historians, linguists
Historic Site
Original location Dresden, Germany Dresden, Germany
Time of origin ~1700–1900 1935–1945
General object partially no longer extant tangible object (partially) no longer extant tangible object
properties
Type of object buildings, landscapes, city buildings (3D city model)
Research context
Research function visualization, documentation of the history, presenta- documentation of sources and infor-
tion, exhibition mation, visualization, conceptualization of
building functions, contextualization (city
model), process investigation (change of
use of buildings and devastation due to
WWII)
Table 5. Application of the proposed classification scheme: Project 5, 1st and 2nd Coder
Category Project 5: Freiberg Cathedral – First Coder Project 5: Freiberg Cathedral – Second
(involved) Coder (not involved)
Sources
Type of source remains, terrestrial laser scans, photographs, texts, physical/ material evidence, iconograph-
pictures, plans ical interpretations
Inherent primary source: remaining late Gothic building from primary source
knowledge ~1500 and several other remaining objects, lost
building from 1180
secondary historical sources: pictures, photographs,
plans
secondary contemporary sources: photographs,
plans, texts, architectural investigations by e.g.,
Magirius (1972)
Model quality
Geometrical LOD3 for exterior of both buildings, Simplified very accurate due to 3D scanning technol-
fidelity reconstruction of interior, Detailed modeling of ogy (exact measurements or fidelity data
relevant objects not available)
Radiometric generalized color scheme non color scan
fidelity
Temporal fidelity 9 different models with no temporal changes within one state, no changes
one model
Documentation
Process doc. textual within student assignment further documentation besides text and
images not available on the website
Results documen- not indexed by metadata unknown
tation
Technology
Technological VR modeling based on acquired data and semi-automated genesis on the basis of
domain plans/pictures scan data; manual rework for illustrational
purposes.
Model genetic human-driven modeling of obstacles, semi- unknown
workflow automated for triangulation of point clouds (ceiling,
Goldene Pforte [Golden Portal]), photo texture for
ceiling
Tools RiScanPRO, FARO SCENE, Maxon Cinema 4D most probably polygon model based on
point cloud.
Geometry data polygon models semi-automated genesis on the basis of
type scan data; manual rework for illustrational
purposes.
Project Cooperation
Project scale ~400 man hours for acquisition, modeling and output 3 university professors, unknown number
of technical professionals and a group of
students
Project dates 2014 April 1, 2014 – September 30, 2014
Involved Compe- art history, linguistics, geodesy, cartography art history, German studies, geodesy,
tencies cartography, 3D design
Historic Site
Original Location Freiberg, Germany Freiberg, Saxony, Germany
Time of origin 1180–2014 around 1225-1230, deconstructed and
rebuilt at the end of 15th century
General object (partially) no longer extant tangible object tangible object
properties
Type of object building, building parts and smaller objects (e.g., sculpting and stonemasonry, sandstone,
pulpit) original color lost
Research context
Research function visualization, conceptualization of functions visualization, illustration of iconography
and architectural concepts process inves-
tigation (change of use of buildings and
devastation due to WWII)
Institute of Art and Music Studies and the Media Center of the Technische Universität
Dresden with the Chair for German Literature and Medieval Studies of the Technische
Universität Chemnitz. The project took place in 2014 with 30 student participants at
both bachelor and master level. The project included students in the fields of art history,
linguistics, and geoscience. The humanities students researched aspects of architecture
within the cathedral and its purpose for communication with the visitors. Visualizations
by the geoscience students, such as reconstructions and animations, illustrated the
results.
4 Implications
The experience with the applications of the proposed scheme for projects [cf. 8] was as
follows: All testers shared the overall assessment that application was easy and the
categories were easily distinguishable. Comprehension problems of specific criteria,
however, were also recorded: For example, a “technological domain” could not always
be determined due to its non-accordance to any specific domain. Likewise, the category
“inherent knowledge” was not intuitively applicable, and the testers could only identify
the requested concepts after consulting the provided examples. In a broader perspective
some general problems seem to result from different meanings of technical terms and
concepts within different disciplines, for example the term level of detail (LOD) [25].
Distinction between primary and secondary sources of information makes more sense
since reconstructions also comprise projects which deal with complex time layers and
variable cultural impacts on the objects of study. Once temporal change is proved or at
least likely, primary sources have to be combined with secondary sources to meet the
demand for comprehensive final information. As evaluated practically, satisfying
completion of the table will only be likely for people very familiar with the project. It
cannot be done based solely on publications and project records as many of the
requested aspects are not reflected in documentations or reports. In other words, project
outsiders looking at an overview of reconstructions with a certain research focus may
find it difficult to arrive at a comprehensive and consistent set of application tables. As
is predictable, categories with discrete values such as date or location lead to perfectly
comparable results. Supposedly due to the provided anchor examples, some other
categories were also interpreted similarly by the encoders. The project scale was—with
one exception—generally described as an amount of person-hours, object types by
object classes (e.g., buildings or landscapes), and available competence, indicated by
the professional background of the staff. Geometric fidelity was mostly quantified using
the LOD scheme, but nonetheless interpreted heterogeneously. Qualitative categories,
for example “inherent knowledge”, were partly misinterpreted, having been filled with
duplicates of the preceding category “sources”. Another category with highly hetero-
geneous interpretation was “research function”. For projects 2 and 5 we had tested
coding by both insiders and external individuals. It was found that 80–90 % of cate-
gories could be clarified without direct project involvement. While descriptions by
external persons were rarely incorrect, both quality and level of detail fell significantly
below that of the description by insiders. Unsurprisingly, the retrieval of information on
the project scale from documentation alone was fairly limited. Information on research
function and inherent knowledge may be derived from project descriptions, but is not
comprehensive or sufficient in many cases.
5 Conclusion
Some general remarks can be made regarding application of the proposed scheme. All
involved coders assessed the scheme as easily applicable. It reportedly took 10–15 min
to fill in all the categories. Sufficient completion required insider knowledge of the
project, but approximately 80–90 % of categories could also be completed—in reduced
detail and complexity—by outsiders. With regards to the reliability of coding,
approximately one third of all categories required discrete values like time, location, or
object classes, leading to fully comparable results. For another major group of cate-
gories, coders employed well-established classification methods, such as staff disci-
plinary background to describe involved competencies, media types to classify sources,
or LOD to distinguish between various geometric fidelities. One shortcoming is that
schemes such as the LOD classification were not interpreted identically by all coders
and should be further explained by anchor examples. Finally, some qualitative cate-
gories like research function or inherent knowledge were handled highly heteroge-
neously and may require either further standardization or reclassification as optional.
Another important remark from one of the testers was that spatial scale, a key factor
which determines all reconstruction processes, was not included in our scheme. While
we tested for the quality of the output, we did not focus on usage. As stated before, no
alternative metadata scheme should be proposed as an outcome of our research. In fact,
the intention was to identify generic properties of digital 3D reconstruction projects.
The aim was to provide a determining framework to assist investigations, for example
into the evolution of a disciplinary identity or the historical development of digital 3D
reconstruction. Moreover, the scheme might provide guidance for methodological
assessment and discourse. Nevertheless a future task would be to synchronize with
extant reference ontologies, especially as many elements for describing projects in
general and objects in particular are highly elaborated within these schemes. Therefore,
we encourage researchers to adapt the scheme according to their needs. Moreover, we
would like to invite other researchers to assess, amend, extend, and further develop the
proposed core scheme.
Acknowledgements. The research activity described in this paper was funded by the German
Federal Ministry of Education and Research (no. 01UG1520).
References
1. Krishnan, A.: What are Academic Disciplines. Some Observations on the Disciplinarity
vs. Interdisciplinarity Debate. National Centre for Research Methods, University of
Southampton, Southampton (2009)
2. Wengenroth, U.: Was ist Technikgeschichte? (1998)
3. Beacham, R., Denard, H., Niccolucci, F.: An introduction to the London charter. In:
Ioannides, M., Arnold, D., Niccolucci, F., Mania, K. (eds.) Papers from the Joint Event
CIPA/VAST/EG/EuroMed Event, pp. 263–269 (2006)
4. Bendicho, V.M.L.-M.: The principles of the Seville Charter. In: Proceedings of XXIII CIPA
Symposium, Prague, Czech Republic, 12–16 September 2011 (2011)
5. Sürül, A., Özen, H., Tutkun, M.: ICOMOS digital database of the cultural heritage of
Trabzon. In: Proceedings of XIX CIPA Symposium, Antalya, Turkey, 30 September–
4 October 2003 (2003)
6. Kiouss, A., Karoglou, M., Labropoulos, K., Moropoulou, A., Zarnic, R.: Recommendations
and strategies for the establishment of a guideline for monument documentation harmonized
with the existing European standards and codes. In: Proceedings of XXIII CIPA
Symposium, Prague, Czech Republic, 12–16 September 2011 (2011)
7. Pfarr, M.: Dokumentationssystem für Digitale Rekonstruktionen am Beispiel der Grabanlage
Zhaoling, Provinz Shaanxi, China, Ph.D. Thesis, Darmstadt (2009)
8. Münster, S., Köhler, T.: 3D reconstruction of cultural heritage artifacts. A literature based
survey of recent projects and workflows. In: Virtual Palaces, Part II. Lost Palaces and their
Afterlife. Virtual Reconstruction between Science and the Media, München (2016)
9. Doerr, M.: The CIDOC conceptual reference module. An ontological approach to semantic
interoperability of metadata. AI Mag. 24(3), 75–92 (2003)
10. Felicetti, A., Lorenzini, M.: Metadata and tools for integration and preservation of cultural
heritage 3D information. In: Proceeding of XXIII CIPA Symposium, Prague, Czech
Republic, 12–16 September 2011 (2011)
11. Ronzino, P., Niccolucci, F., D’Andrea, A.: Built heritage metadate schemas and the
integration of architectural datasets using CIDOC-CRM. In: Boriani, M., Gabaglio, R.,
Gulotta, D. (eds.) Online Proceedings of the Conference BUILT HERITAGE 2013
Monitoring Conservation and Management, Milano, pp. 883–889 (2013)
12. Pfarr-Harfst, M.: Documentation system for digital reconstructions. Reference to the
Mausoleum of the Tang-Dynastie at Zhaoling, in Shaanxi Province, China. In: 16th
International Conference on “Cultural Heritage and New Technologies”, Vienna, 2011,
pp. 648–658 (2011)
13. Niccolucci, F.: Setting standards for 3D visualization of cultural heritage in Europe and
beyond. In: Bentkowska-Kafel, A., Denard, H., Baker, D. (eds.) Paradata and Transparency
in Virtual Heritage, pp. 23–36. Ashgate, Burlington (2012)
14. Rasch, R.F.R.: The nature of taxonomy. Image 19(3), 147–149 (1987)
15. Bowker, G.C., Star, S.L.: Sorting Things Out: Classification and Its Consequences. MIT
Press, Cambridge (1999)
16. Huvila, I.: Sorting the metaverse out and how metaverse is sorting us out. In: Power, D.,
Teigland, R. (eds.) The Immersive Internet: Reflections on the Entangling of the Virtual with
Society, Politics and the Economy, pp. 192–203. Palgrave MacMillan, Basingstoke (2013)
17. Münster, S., Kuroczyński, P., Pfarr-Harfst, M., Grellert, M., Lengyel, D.: Future research
challenges for a computer-based interpretative 3D reconstruction of cultural heritage – a
German community’s view. In: ISPRS Annals of the Photogrammetry, Remote Sensing and
Spatial Information Sciences (XXV International CIPA Symposium) II-5-W3, pp. 207–213
(2015)
18. Münster, S., Hegel, W., Kröber, C.: A classification model for digital reconstruction in
context of humanities research. In: Münster, S., Pfarr-Harfst, M., Kuroczyński, P., Ioannides,
M. (eds.) How to Manage Data and Knowledge Related to Interpretative Digital 3D
Reconstructions of Cultural Heritage? LNCS. Springer, Cham (in print)
19. Mayring, P.: Qualitative Content Analysis. Forum Qualitative Sozialforschung 1(2), Art.
20 (2000)
20. Münster, S., Ioannides, M.: The scientific community of digital heritage in time and space.
In: Guidi, G., Scopigno, R., Torres, J.C., Graf, H. (eds.) 2nd International Congress on
Digital Heritage 2015, Granada (2015). 978-1-5090-0048-7/15
21. Kelle, U.: The developement of categories different Documentation system for digital
approaches in grounded theory. In: Bryant, A., Charmaz, K. (eds.) The SAGE Handbook of
Grounded Theory, pp. 191–213. SAGE (2010)
22. Lombard, M., Snyder-Duch, J., Bracken, C.C.: Content analysis in mass communication:
assessment and reporting of intercoder reliability. Hum. Commun. Res. 28, 587–604 (2002)
23. Lamnek, S.: Qualitative Sozialforschung. Lehrbuch, Weinheim (2005)
24. Bourgeois, J., De Wulf, A., Goosens, R., Gheyle, W.: Saving the frozen Scythian tombs of
the Altai Mountains (Central Asia). World Archaeol. 39, 458–474 (2007)
25. Luebke, D., Watson, B., Cohen, J.D., Reddy, M., Varshney, A.: Level of Detail for 3D
Graphics. Elsevier Science Inc., Atlanta (2002)
Digital Preservation of Cultural Heritage: Balinese
Kulkul Artefact and Practices
Cokorda Pramartha ✉ and Joseph G. Davis

( )
School of Information Technologies, The University of Sydney, Sydney, Australia

{cokorda.rai,joseph.davis}@sydney.edu.au
Abstract. One of the goals of digital preservation of cultural heritage is to gather,

refine, maintain, and share cultural resources that can subsequently be used and
developed by scholars, members of the community, and future generations. We
present the details of our research dealing with one aspect of Balinese culture, the
Balinese traditional communication system (kulkul), undertaken in the Indonesian
island of Bali. We introduce a new framework based on Balinese cultural prin‐
ciples (Tri Hita Karana and Desa Kala Patra) to capture, classify, and organize
cultural artefact and practice knowledge, and design and develop an online digital
portal prototype to enable the sharing and growth of knowledge related to the
Balinese kulkul. This knowledge is held largely in tacit form in the Balinese
community, poorly documented, and fragmented, which makes the preservation
difficult and yet crucial. The aim of the project is to document, preserve, and
educate the Balinese community and the younger generations in particular on an
important aspect of Balinese culture. This community will be encouraged not only
to learn about kulkul and related practices but also contribute their own knowledge
to enable the online digital portal to evolve into a living repository of Balinese
cultural knowledge. The basic kulkul knowledge and understanding was obtained
through in-depth interviews with selected Balinese cultural experts and knowl‐
edgeable community members (Professors from a Balinese University, spiritual
leaders, senior community leaders, and craftsmen). As part of the digital portal,
our project also includes the development of a basic ontology of key kulkul-related
concepts and terms, and their inter-relationships to support the semantic searching
and browsing of online resources.
Keywords: Balinese kulkul · Digital heritage · Digital portal · Ontology · Living

cultural repository
1 Introduction
Over the past few decades, digitization of cultural heritage and natural history has gained
much attention from researchers, practitioners, and memory institutions (galleries, libra‐
ries, archives, museums, and natural history institutions). Digital preservation of cultural
heritage is a complex and interdisciplinary task involving computer science, history,
library, literature, information science and art. One of the goals of digital preservation
is gathering, refining, maintaining, and sharing the cultural resources that subsequently

DOI: 10.1007/978-3-319-48496-9_38
492 C. Pramartha and J.G. Davis
can be used and developed by scholars, members of the community, and younger gener‐
ations. Moreover, digitization is used for creating new means of accessing cultural
information, where the users and/or future generations have the ability to learn, under‐
stand, and further develop the digital resources through the Internet.
Many countries all around the world face the problem of cultural heritage extinction
or depletion, and many of the objects and cultural practices are poorly documented,
largely tacit, and fragmented [1–4]. These contribute in making the preservation project
difficult, and yet crucial. The fragmentation of cultural knowledge presents a major
challenge for cultural preservation, rendering it time-consuming and labor-intensive.
Technological developments such as the Internet, smartphones, and advances in IT allow
for a unique and crowd driven solution to this problem. By undertaking a crowd driven,
systematic approach to gather, store, check and organize cultural information, we aim
to integrate the diverse knowledge, make it widely available through appropriate frame‐
work and platforms, and test the accuracy and validity of our approach towards
preserving and extending cultural heritage.
1.1 Aims and Significance

This project aims to enhance the Balinese community understanding and the younger
generations in particular on an important aspect of Balinese culture. Much of this
knowledge is deeply tacit and collectively held within the Balinese community. This is
the first stage of our digital project which externalizes the complex body of knowledge
and makes it available through an online digital portal for the benefit of the Balinese
community as well as the large number of people with deep interest in Balinese culture
and traditions.
The objective of this paper is to propose a knowledge organization and classification
framework for Balinese cultural heritage, and to design and develop an online digital
portal prototype system for preserving, archiving, and sharing knowledge related to
unique Balinese culture artefacts by tapping into the community-level understanding
and knowledge of cultural heritage.
1.2 Research Problem
A problem encountered in cultural digitization projects is the preservation of information

beyond the actual object, such as contextual and cultural practice-related information.
UNESCO identified such practices as very fragile by their very nature and easily
forgotten, making the digitization of this cultural information is vital for preservation.
Bali is renowned of worldwide tourist destination because of its highly developed
culture such as art, traditional dance, painting, music, language, etc. The information
relating to the Balinese cultural knowledge is complex. Even for cultural and community
experts it is far from straightforward to model this domain of knowledge. Much of this
knowledge is held largely in tacit form and rooted in the Balinese community. Further‐
more, without clear documentation of Balinese culture [5, 6] specifically Balinese kulkul,
makes this study even more challenging and significant. Also, the fact that kulkul prac‐
tices are diverse and not standardized makes it particularly important to understand and
Digital Preservation of Cultural Heritage 493
capture the information as much as possible in the right context, and putting it together
in the right form of an online digital portal.
2 Background and Related Literature
2.1 Bali and Balinese Culture
The island of Bali is one of thousands island, and one of the smallest provinces of the
Indonesia archipelago. The Balinese culture is maintained through daily practice and
religious rituals that keeps the traditions of the past alive [5]. This tradition passes from
one generation to another by letting the next generation learns through experience every
process that related to the local Balinese culture.
The Balinese culture built on the top Balinese traditional community such as custom
village (desa adat or desa pekraman), banjar, and sekaa. Religious ceremonies which
are integral thinking and attitudes of the Balinese, continue as frequently and as impor‐
tantly as ever, and remain relatively unchanged over the years in spite of modernization
trends [7].
Kenthongan is an ancient traditional communication tool which is made of bronze
with elongated holes [8], which is widely known in the Indonesian archipelago. In Bali
this kethongan is known as kulkul (Fig. 1) that is owned by every traditional Balinese
community (desa adat, banjar adat, and sekaa). Kulkuls are made of wood or bamboo
and are installed in the Bale kulkul (typically near temples, village, and banjars) in every
village. These objects, the cultural practices, and messages surrounding the different
kulkul sounds represent a distinct and unique dimension of Balinese cultural heritage.
The kulkul acts as an alarm in the community for ceremonies and hazard e.g. gatherings
the community, fire, flood, murder, etc. [9], and it vary from one Balinese village to
another.
Fig. 1. Balinese kulkul
2.2 Digital Preservation Within Cultural Heritage
Cultural heritage tends to be seen as the product of the physical cultural traditions and
spiritual achievements in the form of the value of the past. Turning towards the digital
age, cultural heritage institutions such as Galleries, Archives, Museums and Libraries
(GLAM) started to identify the need for digitizing their various collections and making
them available online [10].
There are many reasons why cultural heritage collections should be available online.
The most important purpose is to make legacy cultural knowledge accessible, further
developed, and passed on to future generations. Also, this will allow students,
researchers, teachers, and the public to explore and connect with their past. These
cultural heritage collections include sites, objects and intangible things that have
cultural, historical, aesthetic, archaeological, scientific, ethnological or anthropological
value to groups and individuals [11]. Several digital works and project initiatives have
been carried out to prevent cultural knowledge erosion in Indonesia [1, 3, 12].
3 Methodology
This study is based on design science research methodology (DSRM) [13]. It will assess
the ability of experts’ and non-experts’ contributions to preserve, and experiment with
the prototype system to be created. The DSRM (Fig. 2) consist of several stages: (1)
Problem identification and motivation; (2) Objectives for solutions; (3) Design and
development; (4) Demonstration and Evaluation; and (5) Communication.
Fig. 2. Design Science Research Methodology (DSRM)
This research employs a problem-centered approach in which the research entry

point is the problem identification and motivation activity. We proceed in this sequence
because of the idea for the research resulted from observation of the problem. The system
design and development on this project (Fig. 3) is using prototyping method and involves
three activities: 1. Cultural experts interview, 2. Prototyping, and 3. Implementation and
testing the online digital portal.
Fig. 3. System Development Life Cycle

3.1 Externalization of Knowledge

The first part of this study involved exploring and gathering knowledge from Balinese
cultural heritage experts and knowledgeable community members about the Balinese
traditional communication tool (Balinese kulkul). This exploratory phase of our research
project is to externalized the knowledge, help to design an online computer web-based
digitization system, and create a basic model of ontology that can be utilized by
computer-based system to facilitate the browsing and searching for information included
in the online digital portal.
The kulkul community knowledge exists in somewhat tacit form, and this knowledge
is shared through socialization (tacit to tacit) of Balinese community social systems from
generation to generation. Tacit knowledge is a knowledge that based on individual
experience, deeply embedded, hard to express and explain, not yet articulated, and equals
as practical know-how knowledge. Thus, one way to represent it is through metaphors,
drawings, and other method that not require any formal language [14]. Furthermore, the
externalization of tacit knowledge is one of the most challenging steps of knowledge
conversion [14–16].
Over a period of two months prior to the interviews, an extensive review of the
literature was performed to help frame the interview questions and explore the key issues
in the context of Balinese kulkul. Interviews using snowball strategy has been chosen
for the method on externalization of this community tacit knowledge. The snowball
strategy is a form of purposeful sampling when the researchers asks participants to
recommend other individuals to the study [17]. This method is appropriate for our project
when we were initially unfamiliar with the topic and the complexity of the research
central phenomenon. The sampling process initiated by an independent expert in each
group exclusively triggers getting the different snowballs rolling. We stopped the snow‐
ball rolling whenever the group of participants repeats the same answers corresponding
to the questions.
Semi-structured interview approach was selected as the means for information
collection. This format of interview allows for specific questions and topics to be
addressed while offering the respondents an opportunity to give additional feedback and
elaborate further on any aspect of their experience they considered relevant to this study
[18]. The semi-structured interview questions consisted of 23 questions, and fall into
three main categories: (a) history of kulkul, (b) kulkul artefact, and (c) kulkul practices.
The interviews took place during the period of 03rd of August to the 12th of November
2015. In order to collect accurate data and to enable verbatim transcription, they were
audio-recorded, always with the interviewee’s consent. The interviews used a combi‐
nation of Bahasa and Balinese language. The participants’ age was between 47 – 65
years old, their average experience in their field is more than 10 years, all participants
were male. They came from the west, east, south, and central regions of Bali.
The basics of kulkul system was collected incrementally from four groups of experts.
These groups of interviewees consisted of five Professors from a Balinese University
(who are acknowledged experts in methods of traditional communication), four Balinese
spiritual leaders (who are identified because they have the knowledge about the spiritual
significance of these traditions), eight senior community leaders in Bali (identified as
key leaders who are the person in charge to ring the kulkul in banjar or village), and two
kulkul craftsmen (identified because they possess practical knowledge in making the
kulkul artefact).
Each interview session took for about 30 min up to 120 min. Many of the respondents
had difficulties in expressing the knowledge surround the kulkul domain. They under‐
stood how the kulkul is practiced in the community, however it was sometimes hard for
them to express it in the formal language. Therefore, occasionally researchers had to
repeat the interview due to the complex and relatively tacit knowledge related to the
kulkul domain.
We used the knowledge externalization method proposed by Nonaka and Takeuchi
Fig. 4. First, metaphor is a method to express what the interviewees know, but hard to
say by asking them to create a symbol and synthesis to find the distance between
symbols. This method is used to merge two different and distant areas of experience into
single image or understanding. Second, analogy is used to reconcile contradictions and
make distinctions between symbols. Finally, the knowledge was crystalized into a model
by constructing the basic kulkul domain ontology.
Fig. 4. Externalization knowledge methodology
During the interview we asked some of the interviewees to demonstrate how they
practice the kulkul in the community to help us understand the different practices of
kulkul. For example, two of the interviewees literally said that they practice two frames
(duang tulud bulus) for the situation of hazard, nevertheless the way they sounded the
kulkul are completely different (the total number of beat and the rhythm) as shown on
Fig. 5.
Fig. 5. Two frames of kulkul bulus at (a) Gianyar region, and (b) Karangasem region
3.2 Knowledge Classification and Organization
Our digital portal framework (Fig. 6) to capture, classify, and organize the richness of
kulkul knowledge is based on the outcome of the interviews which enabled us to elab‐
orate on the key Balinese cultural principles (Tri Hita Karana and Desa Kala Patra) as
they relate to kulkul. Tri Hita Karana is the Balinese belief or philosophy of balance
and harmony [19–21] including three factors: (1) Universal (parahyangan), (2) envi‐
ronment (palemahan), which refers to harmony with nature, and (3) people
(pawongan), which implies harmony among people. This first set of concepts is related
to the kulkul practices where kulkul sound is a part of every Balinese religious ceremony
(parahyangan) that is mainly held in the temple area. Also, this sound is used for any
type of activities that involve a large number of people in the Balinese traditional
community (pawongan). The second set of concepts called the space (desa), time (kala),
and circumstances (patra) [21] dictates that humans are not passive individuals but who
will always try to adapt themselves to their environment that influences their lives. The
Desa Kala Patra concept can be used to explain the variations in kulkul practices in the
Balinese community.
Fig. 6. The Balinese kulkul framework
The kulkul knowledge classification cannot separate the interdependence between

the artifact (tangible) and its practices (intangible). Also, this kind of cultural heritage
classification systems are still being debated by the UNESCO on the Preamble of the
2003 Convention for the Safeguarding of the Intangible Cultural Heritage [22].
Information can be viewed at two different levels: syntactic where this information
based on the volumes of the information, and semantic more focus on the meaning of
the information and how it related to the other concept. Ontology is a formal, explicit
specification of a shared conceptualization [23], and have been developed to provide a
machine-processable semantics of information sources that can be communicated
between different agents (software and humans) [24]. Moreover, ontology is one of
alternatives to representing the domain of knowledge, where this method helps us to do
semantic modeling of concepts, and the axioms in ontology are represented in logic
languages [25].
The CIDOC CRM1 is a well-known cultural ontology that enables information inte‐
gration, mediation and interchange of heterogeneous cultural heritage data and their
correlation with digital library and archive information, and this cultural ontology has
been studied for this project. The CIDOC-CRM has been used to abstract hundreds of
schemata (80 classes and 130 relationships) in various museum disciplines, and helps
improve the semantic of knowledge from distributed database of cultural heritage
1
http://www.cidoc-crm.org.
[26–28]. However, less than five percent of its concept has been used by museum [29],
and the CIDOC-CRM is too museum centric [30].
The ontology development falls into two categories: (i) developed from scratch [31],
and (ii) reusing and integrating with an existing ontology [32] based on the knowledge
representation needed. In this project we develop the ontology from scratch, and the
purpose of applying ontological approach is to enable flexibility of the knowledge
growth, as the collecting of kulkul knowledge was done incrementally. Therefore, a
formal reasoning can be applied when this knowledge evolves.
4 Prototyping, Design, and Implementation of Digital Portal
The online digital portal prototype is currently under development. The development of
basic kulkul ontology (class, object properties, and data properties) Fig. 7 used a
protégé2 ontology editor, and the OWL2 language. This kulkul ontology serves as back‐
bone to support semantic browsing and searching facilities of our system. The apache
Jena3 framework was chosen as ontology triple store for our development, while
Fuseki4 is used as a middleware layer to interfacing our web application to the ontology.
In addition to that, the EasyRdf5 API was used to provide a PHP API to Fuseki. The
browsing results are retrieved and displayed using JSON and AJAX calls. Our applica‐
tion supports the following facilities:
– Populate: allows multiple online users to contribute by populating more details of
Balinese kulkul.
– Browsing: allows users to browse through different part of the ontology by navigating
through kulkul classification hierarchies.
– Searching: allows users to construct a query relating to one or more attributes of the
kulkul artifact and practices as inputs and displaying a selected attribute for the output.
Fig. 7. Kulkul ontology
2
http://protege.stanford.edu/.
3
http://jena.apache.org/.
4
https://jena.apache.org/documentation/serving_data/.
5
http://www.easyrdf.org/.
We have presented the details of our research dealing with one aspect of Balinese culture,
the Balinese traditional communication system (kulkul). Our contributions include
externalizing the largely tacit and fragmented kulkul knowledge, and developing an
ontology to represent this knowledge. The kulkul framework based on Balinese cultural
principles (Tri Hita Karana and Desa Kala Patra) is to capture, classify, and organize
cultural artefact and practice knowledge, and design and develop an online digital portal
prototype to enable the sharing and growth of knowledge related to the Balinese kulkul.
We are currently working on analyzing and testing the usability of our online digital
portal system by having community-based crowdsourcing to contribute more details in
order to refine and update community cultural resources. The users’ feedback will be
used as inputs to expand the kulkul ontology and to enhance the digital portal.
Acknowledgement. This project received funding from The Indonesia Endowment Fund for
Education (LPDP) grant no. PRJ-02/LPDP/2013 and Competitive Research Grant Ristekdikti
grant no. 486.24/UN14.2/PNL.01.03.00/2016.
References
1. Sanabila, H.R., Manurung, R.: Towards automatic wayang ontology construction using
relation extraction from free text. In: EACL 2014, p. 128 (2014)
2. Rosner, D., Roccetti, M., Marfia, G.: The digitization of cultural practices. Commun. ACM
57, 82–87 (2014)
3. Hasibuan, Z.A.: An overview of integrated approach to digital preservation: case study of
Indonesian e-Cultural heritage and natural history information retrieval system. In:
International Conference on Advanced Computer Science and Information System
(ICACSIS), pp. 31–36 (2011)
4. Hui, D., Siwei, Y., Ying, J.: Knowledge representation of Chinese genealogical record of VIPs
in KMT and CPC. In: Ninth International Conference on Hybrid Intelligent Systems, HIS
2009, vol. 3, pp. 116–120 (2009)
5. Walker, B., Helmi, R.: Bali Style. Thames and Hudson, London (1995)
6. Covarrubias, M.: Island of Bali. Periplus Editions (HK) Limited, Singapore (2008)
7. Suryani, L.K., Jensen, G.D.: Trance and Possession in Bali: A Window on Western Multiple
Personality, Possession Disorder, and Suicide. Penerbit ITB, Bandung (1999)
8. Kebudayaan, D.P.: Ensklopedia Musik Indonesia. Departemen Pendidikan dan Kebudayaan,
Proyek Inventarisasi, dan Dokumentasi Kebudayaan Daerah (1985)
9. Goris, D.R., Dronkers, D.P.L.: Bali Atlas Kebudayaan. Pemerintah Republik Indonesia (1955)
10. Zlodi, G., Ivanjko, T.: Crowdsourcing digital cultural heritage. izlazi u samo elektroničkom
izdanju: NE (2013)
11. http://www.unesco.org/new/en/brasilia/culture/world-heritage/heritage-legacy-from-past-
to-the-future/
12. Pramartha, C.R.A., Dwidasmara, I.B.G.: The composition approach non-QWERTY keyboard
for Balinese script. In: 2014 IEEE Canada International Humanitarian Technology
Conference - (IHTC), pp. 1–4 (2014)
13. Peffers, K., Tuunanen, T., Rothenberger, M.A., Chatterjee, S.: A design science research
methodology for information systems research. J. Manage. Inf. Syst. 24, 45–77 (2007)
14. Koskinen, K.U., Pihlanto, P., Vanharanta, H.: Tacit knowledge acquisition and sharing in a
project work context. Int. J. Proj. Manage. 21, 281–290 (2003)
15. Abidi, S.S.R., Cheah, Y.N., Curran, J.: A knowledge creation info-structure to acquire and
crystallize the tacit knowledge of health-care experts. IEEE Trans. Inf. Technol. Biomed. 9,
193–204 (2005)
16. Nonaka, I., Takeuchi, H.: The Knowledge-Creating Company: How Japanese Companies
Create the Dynamics of Innovation. Oxford University Press, Now York (1995)
17. Creswell, J.W.: Educational Research: Planning, Conducting, and Evaluating Quantitative
and Qualitative Research. Merrill, Upper Saddle River (2005)
18. Louise Barriball, K., While, A.: Collecting data using a semi-structured interview: a discussion
paper. J. Adv. Nurs. 19, 328–335 (1994)
19. Jensen, G.D., Suryani, L.K.: The Balinese People: A Reinvestigation of Character. Oxford
University Press, Singapore (1992)
20. Windia, W., Pusposutardjo, S., Sutawan, N., Sudira, P., Supadmo Arif, S.: Transformasi
sistem irigasi subak yang berlandaskan konsep TRI Hita Karana. SOCA (Socio-Economic of
Agriculturre and Agribusiness) 5 (2005)
21. Pitana, I.: Tri Hita Karana – the local wisdom of the balinese in managing development. In:
Conrady, R., Buck, M. (eds.) Trends and Issues in Global Tourism 2010, pp. 139–150.
22. http://www.unesco.org/culture/ich/en/convention
23. Gruber, T.R.: A translation approach to portable ontology specifications. Knowl. Acquisition
5, 199–220 (1993)
24. Fensel, D.: Ontologies: A Silver Bullet for Knowledge Management and Electronic
Commerce. Springer, Heidelberg (2004)
25. Martinez-Cruz, C., Blanco, I.J., Vila, M.A.: Ontologies versus relational databases: are they
so different? A comparison. Artif. Intell. Rev. 38, 271–290 (2012)
26. Hong-Zhe, L., Bao, H., Jing, W., Jun-Kang, F.: An information flow based approach to
semantic integration of distributed digital museums. In: International Conference on Machine
Learning and Cybernetics, pp. 4430–4437 (2006)
27. Guoxin, T., Tinglei, H., Zheng, Z.: A knowledge modeling framework for intangible cultural
heritage based on ontology. In: Second International Symposium on Knowledge Acquisition
and Modeling, KAM 2009, vol. 1, pp. 304–307 (2009)
28. Hong-Zhe, L.: Global ontology construction for heterogeneous digital museums. In:
International Conference on Machine Learning and Cybernetics, vol. 7, pp. 4015–4019 (2007)
29. Doerr, M., Iorizzo, D.: The dream of a global knowledge network: a new approach. J. Comput.
Cult. Herit. 1, 1–23 (2008)
30. Brownlow, R., Capuzzi, S., Helmer, S., Martins, L., Normann, I., Poulovassilis, A.: An
ontological approach to creating an andean weaving knowledge base. J. Comput. Cult. Herit.
8, 1–31 (2015)
31. Cristani, M., Cuel, R.: A survey on ontology creation methodologies. Int. J. Seman. Web Inf.
Syst. (IJSWIS) 1, 49–69 (2005)
32. Gómez-Pérez, A., Rojas-Amaya, M.D.: Ontological reengineering for reuse. In: Fensel, D.,
Studer, R. (eds.) EKAW 1999. LNCS (LNAI), vol. 1621, pp. 139–156. Springer, Heidelberg
(1999). doi:10.1007/3-540-48775-1_9
Interconnecting Objects, Visitors, Sites
and (Hi)Stories Across Cultural and Historical
Concepts: The CrossCult Project
Costas Vassilakis1 , Angeliki Antoniou1 , George Lepouras1 ,

Manolis Wallace2(B) , Ioanna Lykourentzou3 , and Yannick Naudet3
1
Human-Computer Interaction and Virtual Reality Lab,
Department of Informatics and Telecommunications,
University of the Peloponnese, 22 131 Tripolis, Greece
{costas,angelant,gl}@uop.gr
2
Knowledge and Uncertainty Research Laboratory,
Department of Informatics and Telecommunications,
University of the Peloponnese, 22 131 Tripolis, Greece
[email protected]
3
Luxembourg Institute of Science and Technology,
5 avenue des Hauts-Fourneaux, 4363 Esch/Alzette, Luxembourg
{ioanna.lykourentzou,yannick.naudet}@list.lu
http://hci-vr.dit.uop.gr/, http://gav.uop.gr, http://www.list.lu/
Abstract. Human History, is a huge mesh of interrelated facts and

concepts, spanning beyond borders, encompassing global aspects and
finally constituting a shared, global experience. This is especially the case
regarding European history, which is highly interconnected by nature;
however, most History-related experiences that are today offered to the
greater public, from schools to museums, are siloed. The CrossCult
project aims to provide the means for offering citizens and cultural venue
visitors a more holistic view of history, in the light of cross-border inter-
connections among pieces of cultural heritage, other citizens viewpoints
and physical venues. To this end, the CrossCult project will built a
comprehensive knowledge base encompassing information and semantic
relationships across cultural information elements, and will provide the
technological means for delivering the contents of this knowledge base
to citizens and venue visitors in a highly personalized manner, creat-
ing narratives for the interactive experiences that maximise situational
curiosity and serendipitous learning. The CrossCult platform will also
exploit the cognitive/emotional profiles of the participants as well as
temporal, spatial and miscellaneous features of context, including holi-
days and anniversaries, social media trending topics and so forth.
Keywords: Adaptation · User profiles · Mobile applications
1 Introduction
“CrossCult: Empowering reuse of digital cultural heritage in context-aware cross-
cuts of European history” is a newly started project, supported by the European

DOI: 10.1007/978-3-319-48496-9 39
502 C. Vassilakis et al.
Union under the H2020-REFLECTIVE-6-2015 “Innovation ecosystems of digital

cultural assets” funding scheme. The CrossCult project aims to help European
citizens understand and perceive their common past and present in an holis-
tic manner, while fostering retention and promoting reflection. To this end, the
CrossCult project adopts guidelines formulated in recent research [5,6] regarding
the design of the interactive experiences and their narratives; these guidelines
can be codified as four major principles as follows:
– Raise consciousness about the importance of History

– Tackle the study of History from a multi-faceted perspective
– Approach History not only through the written texts from successive eras,
but also through all the traces left by those societies (archaeological remains,
iconography, epigraphy, numismatics, architecture, art, etc.)
– Reckon that there are no absolute truths in History, but various possible inter-
pretations of the archaeological remains and contrasting viewpoints
In this light, the cultural heritage sites, historic events and (digital) cultural
assets are not presented as disconnected but rather they target at increasing
understanding of the interrelated nature of culture and history. In doing so, the
CrossCult project and platform allows the interconnections of objects, resources,
visitors, sites and (hi)stories across European Cultural and Historical concepts
and events. In addition, CrossCult uses an extensive knowledge base, using exist-
ing cultural repositories (e.g. Europeana) and a specially designed technological
platform to maximize situational curiosity, serendipitous learning and reflection
on different historical and social phenomena, like for example the place of women
in societies, the movement of populations historically and today, establishment
of healing practices, etc.
The project is using cutting edge technology to allow meta-history research
from pilot experience with narratives built from cross-border connections and
crosscutting topics. Furthermore, new technologies for smart venues and whole
cities allow the multi-level, cross-repository interconnection of venues and digi-
tal cultural heritage resources. Personalized, adaptive and context-aware experi-
ences are at the heart of the project in attempt to increase the Quality of Visitor
Experience [13].
The project approach will be implemented and validated through a number
of pilots, designed to operate under different parameters (individual vs. col-
laborative participation; authoritative vs. crowdsourced content; narrative vs.
exploratory vs. serendipitous content delivery; intra-venue vs. inter-venue con-
nections; temporal, vs. spatial vs. topic-related connections; physical presence vs.
virtual presence), and will provide valuable insight on the effect that these para-
meters have on cognitive aspects, such as reflection, (re)interpretation, relation
and comparison.
The remaining of this paper is organized as follows: Sect. 2 focuses on the
personalization aspects in CrossCult, considering that personalization is consid-
ered a key feature to achieve the goals listed above and Sect. 3 describes the
CrossCult infrastructure, outlining the architectural modules of the platform
Interconnecting Objects, Visitors, Sites and (Hi)Stories 503
and their functionality. Finally, Sect. 4 concludes the paper and outlines future
research.
2 Personalization in Cultural Heritage

Previous research has shown the importance of personalized applications for cul-
tural heritage since they are found to increase the quality of visiting experience
[1,12]. Personalized applications in cultural heritage are not only increasingly
popular among cultural institutions [3] but visitors also seem to prefer them [7].
Personalized content can target the needs of a diverse audience and allow the
targeted presentation of big amounts of available cultural information, providing
different information to different people [19]. Cultural heritage institutions also
use personalized applications as means to attract visitors. For a detailed survey
on the use of personalization approaches in cultural heritage institutions one
may refer to [4].
Within our project personalization goes beyond the current state of the art,
aiming to appropriately tailor content, functionalities and presentation so as to
foster reflection, (re)interpretation, relation and comparison, taking into account
aspects such as the users cognitive profile, the current group dynamics, non-
typical connections of cultural objects and so forth.
2.1 The Cold Start Problem and the Role of Social Media
However, although personalization in cultural heritage is useful, creating correct
visitor profiles is a rather demanding task basically due to the short duration of
most visits and the fact that most visitors might only visit a specific space only
once. Within these time restrictions, visitor profiles need to be created quickly
and effectively in terms of their appropriateness for the different visitors. The
question raised though is how to start creating these profiles and where to find
the necessary information.
Different approaches have been employed like using heuristic techniques [9],
probability-based algorithms [11] or user prototypes/personas [15]. The problem
is that although visitors enjoy the benefits of personalized applications in cultural
heritage, they are at the same time reluctant at dealing with form-filling activities
and researchers have to become more creative in applying indirect approaches
for the collection of the needed information for the creation of user profiles.
Over the last years, more and more people are using social media for social
interaction, information seeking, pass time, entertainment, relaxation, communi-
catory utility, convenience utility, expression of opinion, information sharing, and
surveillance/knowledge about others. While interacting with these media, people
provide vast amounts of personal information that can be further exploited for
the creation of user profiles.
It has been demonstrated that peoples behavior on social networks reflects
actual personality traits [17]. Therefore social media can be rich sources of infor-
mation to efficiently create user profiles for personalized content in cultural her-
itage, also dealing with the cold start problem.
Recently a European project explored the potential of Facebook in providing

the necessary information for the creation of museum visitor profiles [2]. How-
ever, research is still at the early stages and algorithms are not always reliable
[16]. In this framework, CrossCultCrossCult aims at providing the necessary
infrastructure for the effective elicitation of user profiles using different means,
including social networks.
2.2 Group Adaptation

Although the above may apply for individuals and the elicitation of individual
user profiles, adaptation for group visitors and elicitation of group profiles is a
whole different story. Visitors rarely visit alone [14], yet most cultural heritage
technologies, such as the ones discussed in the previous paragraphs, are made
for individuals.
Early approaches have provided group visitors with a shared device for a
common experience, but in more recent studies although each visitor has her
own device, the experience is cooperative; for example the CHESS system that
uses interactive storytelling for group visitors [18]. Context-aware applications
for small groups have been also discussed [10] and the mobility of visitors has
been also taken into account [8].
2.3 Personalization and Adaptation in CrossCult

The CrossCult platform will address personalization and adaptation by realizing
both item recommendations and path recommendations.
Item recommendations include points of interest, exhibits, or even individ-
ual resources. Content-based recommendation techniques will be used here to
match the candidates for recommendation (POIs, exhibits or resources) against
the profile of the visitor; additionally, matching the candidates for recommenda-
tion against the current context (e.g. semantics of currently viewed collection,
characteristics of exhibits viewed so far within the visit etc.) may provide oppor-
tunities for re-contextualization of entities and increased serendipity. Collabora-
tive filtering techniques will also be employed, to enhance the visitors experience
using information from the visits of other users with similar profiles (e.g. similar
preferences, interests, visit goals and so forth). The recommender system will
exploit the full set of semantic information that will be available for candidates
for recommendation, by accessing the relevant ontological descriptions that are
stored in the CrossCult repository (Fig. 1).
Path recommendations will not only arrange for suggesting routes that
include exhibits that are deemed of high interest to the user, but also arrange
for making the visiting experience meaningful. To this end, path formulation
will strive to lead the visitor through routes in which the exhibits tell a certain
story (e.g. they relate to social aspects of life in antiquity) or allow the visitor to
reinterpret History from the items that s/he will view along the suggested route.
Furthermore, path formulation will take into account real-time data regarding
the location of other visitors, so as to avoid annoyances, such as high waiting
times or overly crowded areas; due to the real-time nature of crowd location
data, routes may be re-planned to adapt to the current situation. Finally, route
planning will consider the visiting style of the visitors, i.e. their usual pattern
of movement into cultural venues as well as constraints on their available or
planned time; an initial estimate for the duration of the visit can be projected
from previous visits, considering both the venue statistics and the individual
visitor statistics, while the user may modify the estimated value.
All recommendation techniques require that user characteristics, interests
and interaction history are recorded in a user profile, so as to match them either
against the data and metadata of recommendation candidates (for content-based
recommendations) or against profiles of other users (for collaborative filtering-
based approaches). Users however may not offer to directly enter information
through profile completion pages since the related for them is high and the
benefits are not always apparent; moreover, for users that have been recently
registered with the system, formulation of successful recommendations will be
inhibited due to the cold start problem. To tackle these issues, techniques for
implicit profile population and interest extraction are employed complementary
to explicit input. Mini-games and quizzes are used to offer an insight to charac-
teristics and interests of the users, while mapping visitors to personas or stereo-
types is used to assist in identifying recommenders (i.e. other people with similar
interests or characteristics, often called nearest neighbors) for collaborative fil-
tering recommendations. Mini-games and quizzes will be delivered through social
network platforms, exploiting their widespread and dissemination dynamics to
maximize the potential of information gathering.
Recommender systems in CrossCult will exploit trending topics information
from social media to allow venue content curators to promote exhibits or collec-
tions relevant to topical issues; this feature will enable venues to take advantage
of subjects publicity, turning it into interest on their content. In this context,
trending topics are read from social media and matched against the metadata
of exhibits and collections; subsequently, the highest matches are presented to
venue content curators who can then choose how (and if) these matches can be
best used to trigger visitor interest and promote the venue content.
Regarding the aspect of group adaptation, the CrossCult project is in line
with recent developments in museum/visitor studies listed earlier and it recog-
nizes the need for the development of collaborative applications targeting group
visitors as well as individuals. For this reason, group profiles will be created
and content will be adapted according to the groups requirements. To this end,
the personalization mechanisms in the CrossCult project include algorithms to
determine whether an individual is visiting alone, or in a group. While this
can be extracted using a direct question, additional technological means will be
exploited to facilitate user profiling, such as near-field communications, identi-
fication of previously paired Bluetooth devices, identification of users that have
an established friend relationship in social networks and so on.
Finally, recommenders in CrossCult will foster the creation of sporadic social
networks by recommending connections among people that co-visit a venue. In
Fig. 1. The CrossCult platform overall architecture
this respect, the profiles of current venue visitors will be analyzed and groups of
people with related interests and matching personalities will be determined; then
suggestions will be forwarded to members of each group to join an online group
through which they can engage in collaborative live discussion, participate in
educational games or even formulate a physical group. These recommendations
will respect user anonymity, and only after explicit consent of all interested
parties the electronic IDs of the users will be exchanged.
We expect that user interaction within the context of sporadic social net-
works will promote the creation of new views on the venue content and the
connections between exhibits, enriching the venues informational content. Some
of the connections are expected to outlive the duration of the visit (e.g. through
an establishment of a friend relation in social networks), supporting post-
visit actions (including discussions, diffusion of information to other social net-
work members and so forth), which could extend the benefits of the visiting
experience.
3 CrossCult Infrastructrure
To accommodate the functionalities listed in the previous section, the CrossCult

platform employs a layered approach targeting all stakeholders in the user expe-
rience delivery process. These include experience designers, venue experts and
curators, system administrators and visitors.
In this process, venue experts and curators undertake the tasks of (a) curat-
ing digital content, providing textual and multimedia descriptions for exhibits
as well as the associated metadata, and (b) establishing semantic links between
exhibits/exhibit elements. Semantic links may be either established directly, e.g.
by linking a painting depicting a loom to a statue of woman under the rela-
tionship uses, or indirectly by linking two items to the same ontology concept
(e.g. two exhibits representing a loom and a horse are linked to the everyday
life ontology concept). Furthermore, semantic links may be established among
exhibits of different venues, or venue exhibits may be linked to resources in
cultural repositories such as Europeana.
Experience designers employ user profiling tools to arrange for user profile
population, mainly through mini-games and quizzes. Experience designers iden-
tify the opportunities for personalization and exploit exhibit data/metadata,
user profile information and data from analysis of previous visits to the venue to
offer a personalized experience to the visitor. Experience designers also arrange
for offering suitable opportunities to visitors to enter comments as well as per-
sonal views on exhibits and their relations; these inputs may then be presented
to subsequent visitors as personal views of previous users or be reviewed and
accepted by venue experts and curators, and be integrated into the venues offi-
cially delivered information. In this respect, user input handling constitutes a
form of crowdsourcing. Experience designers finally cater for choosing data from
external sources, such as trending topics, calendar information (such as holidays
or events), external sources with POI evaluations and so forth, which will be
exploited to promote personalization and foster topicality.
The last category of the venue-side stakeholders is system administrators,
who will provide the necessary computing and communication infrastructure for
the operation of the applications and the storage and retrieval of information.
All venue-side stakeholders perform their tasks from the CrossCult front-end,
which constitutes a powerful and intuitive web-based environment within provid-
ing the above-listed features. The data entered through the front-end are stored
and organized at the CrossCult back-end which hosts the information reposi-
tory, as well as all services needed to support the visitors application runtime
(the back-end is detailed below). Once all required elements have been entered
and configured via the CrossCult front end, the mobile apps which the visitors
will use can be instantiated.
App instantiation is an automated procedure within which the mobile app
code is generated, using predefined code templates and venue-specific options.
Options mainly pertain to the functionalities that will be available to the users
of the mobile app (i.e. venue visitors), such as entering comments and personal
views, formulating sporadic networks, participating in mini-games and quizzes,
sending messages to other users and so forth; in all cases, content (both offi-
cial and user-contributed, such as comments and texts entered by other users)
is retrieved dynamically from the CrossCult back-end. The mobile apps also
arrange for receiving from the back-end recommendations, notifications and mes-
sages from other users and communicating them accordingly to the visitors.
The CrossCult back-end encompasses a repository for the storage, organi-
zation and retrieval of cultural information and pertinent metadata, while it
additionally hosts modules for delivering the needed functionalities to both the
front-end and the mobile apps. In particular, the back-end arranges for the
following tasks (a) storing user profiles and unifying user profile information
obtained from different sources (e.g. explicit input, mini-games, quizzes etc.);
(b) gathering contextual data (including time of day, calendric information such
as holidays, trending topics and related information from the internet); (c) col-
lecting information about the number of visitors within each venue sub-location;
(d) performing semantic reasoning over the ontological representations to deduce
new relationships from the existing ones, augmenting thus the informational con-
tent; new relationships may be also detected among exhibits of different venues
or exhibits of a venue and external cultural information repositories through
metadata matching; (e) using all the previous listed information to generate rec-
ommendations for the users, regarding the exhibits to see, the paths to follow and
venues to visit, and (f) supporting sporadic social network formulation (includ-
ing identification of groups of visitors with common interests and generation
of recommendations for joining sporadic social networks) and communication
between their members,
4 Conclusions
In this paper we have outlined the main vision of the CrossCult project, which
aims to provide the means for offering citizens and cultural venue visitors a
holistic view of history, breaking cultural, local or viewpoint siloes in History-
related experiences. In this respect, the CrossCult project approaches History
as a huge mesh of interrelated facts and concepts, spanning beyond borders,
encompassing global aspects and finally constituting a shared, global experience.
The main instrument to achieving this goal is personalization, through which
narratives will be created for the interactive experiences that maximise situa-
tional curiosity and serendipitous learning. Personalization will exploit informa-
tion in user profiles, either directly entered or more typically implicitly inferred
and a semantically rich network of cultural information items; these items can
be hosted in a single cultural venue, be dispersed among multiple venues or even
be hosted in cultural repositories, such as Europeana. The CrossCult platform
will also exploit the cognitive/emotional profiles of the participants as well as
temporal, spatial and miscellaneous features of context, including holidays and
anniversaries, social media trending topics and so forth. The CrossCult project
will finally promote group exploration and group activities by employing group
adaptation techniques and fostering the creation and operation of sporadic social
networks, which will interconnect people with similar interests.
The CrossCult project is currently at its development stage. The platform
components, which capitalize on the participants existing technologies, are being
built and will be integrated to form the front-end and back-end components.
Within the CrossCult project four pilot applications will be built, each one tar-
geting to a specific category of History-related experiences: pilot 1 targets large
multi-thematic venues; pilot 2 targets a setting with many small, interconnected
venues; pilot 3 targets a single venue that promotes non-typical cultural connec-
tions among exhibits; and pilot 4 targets a setting involving multiple cities, offer-
ing cross-cultural interplay with their past and present. Through these pilots, the
CrossCult approach will be validated and further enhanced with new features,
taking into account the reactions comments and requests of the users.
Acknowledgment. Part of this work has been funded by the CrossCult H2020
project, Grant #693150.
References
1. Antoniou, A., Lepouras, G.: A study to investigate adaptation aspects for museum
learning technologies. J. Comput. Cult. Herit. 3(2), Article 7, October 2010
2. Naudet, Y., Antoniou, A., Lykourentzou, I., Tobias, E., Rompa, J., Lepouras, G.:
Museum personalization based on gaming and cognitive styles: the BLUE experi-
ment. Int. J. Virtual Communities Soc. Netw. 7(2), 1–29 (2015). Special Issue on
Social Media and Networks for Multimedia Content Management
3. Aoidh, E.M., Bertolotto, M., Wilson, D.: Towards dynamic behavior-based profiling
for reducing spatial information overload in map browsing activity. GeoInformatica
16, 409–434 (2012)
4. Ardissono, L., Kufli, T., Petrelli, D.: Personalization in cultural heritage: the road
travelled and the one ahead. User Model. User-Adap. Inter. 22(1–2), 73–99 (2012)
5. Berglund, S., Ekman, J., Deegan-Krause, K., Knuten, T. (eds.): The Handbook of
Political Change in Eastern Europe. Edward Elgar Publishing, Cheltenham (2013)
6. Burton, A.: Ten Design Principles. Duke University Press, Durham (2011)
7. Findlater, L., McGrenere, J.: A comparison of static, adaptive, and adaptable
menus. In: Proceedings of the SIGCHI Conference on Human Factors in Computing
Systems (CHI 2004), pp. 89–96. ACM, New York (2004)
8. Fosh, L., Benford, S., Reeves, S., Koleva, B., Brundell, P.: See me, feel me, touch
me, hear me: trajectories and interpretation in a sculpture garden. In: Proceedings
of the SIGCHI Conference on Human Factors in Computing Systems (CHI 2013),
pp. 149–158. ACM, New York (2013). http://dx.doi.org/10.1145/2470654.2470675
9. Hameed, M.A., Ramachandram, S., Al Jadaan, O.: Information Gain Clustering
Through Prototype-Embedded Genetic K-Mean Algorithm (IGCPGKA): a novel
heuristic approach for personalisation of cold start problem. In: Proceedings of
the Second International Conference on Computational Science, Engineering and
Information Technology (CCSEIT 2012), pp. 390–395. ACM, New York (2012)
10. Kuik, T., Stock, O., Zancanaro, M., Gorfinkel, A., Jbara, S., Kats, S., Sheidin, J.,
Kashtan, N.: A visitor’s guide in an active museum: presentations, communications,
and reection. J. Comput. Cult. Herit. 3(3), Article 11, February 2011. 25 pages,
http://dx.doi.org/10.1145/1921614.1921618
11. Lin, C., Xie, R., Li, L., Huang, Z., Li, T.: PRemiSE: personalized news recommen-
dation via implicit social experts. In: Proceedings of the 21st ACM International
Conference on Information and Knowledge Management (CIKM 2012), pp. 1607–
1611. ACM, New York (2012)
12. Lykourentzou, I., Claude, X., Naudet, Y., Tobias, E., Antoniou, A., Lepouras, G.,
Vassilakis, C.: Improving museum visitors’ quality of experience through intelligent
recommendations: a visiting style-based approach. In: Intelligent Environments
(Workshops), pp. 507–518 (2013)
13. Naudet, Y., Lykourentzou, I., Tobias, E., Antoniou, A., Rompa, J., Lepouras, G.:
Gaming and cognitive profiles for recommendations in museums. In: SMAP 2013,
pp. 67–72 (2013)
14. Petrelli, D., Not, E.: User-centred design of flexible hypermedia for a mobile guide:
reflections on the HyperAudio experience. User Model. User-Adap. Inter. 15(3–4),
303–338 (2005)
15. Roussou, M., Katifori, A., Pujol, L., Vayanou, M., Rennick-Egglestone, S.J.: A life
of their own: museum visitor personas penetrating the design lifecycle of a mobile
experience. In: CHI 2013 Extended Abstracts on Human Factors in Computing
Systems (CHI EA 2013), pp. 547–552. ACM, New York (2013)
16. Theodoridis, T., Papadopoulos, S., Kompatsiaris, Y.: Assessing the reliability of
Facebook user profiling. In: Proceedings of the 24th International Conference on
World Wide Web (WWW 2015 Companion), pp. 129–130. ACM, New York (2015)
17. Tkalcic, M., Chen, L.: Personality and recommender systems. In: Ricci, F., Rokach,
L., Shapira, B. (eds.) Recommender Systems Handbook, pp. 715–739. Springer,
New York (2015)
18. Vayanou, M., Katifori, V., Chrysanthi, A., Antoniou, A.: How to coordinate visitor
actions? In: Proceedings of the 19th ACM Conference on Computer-Supported
Cooperative Work and Social Computing, San Francisco, 27 February–2 March
2016
19. Walczak, K., Wojciechowski, R., Cellary, W.: Dynamic interactive VR network
services for education. In: Proceedings of the ACM Symposium on Virtual Reality
Software and Technology (VRST 2006), pp. 277–286. ACM, New York (2006)
Project Paper: 3D Reconstruction and
3D Modelling
Debate and Considerations on Using Videos for Cultural
Heritage from Social Media for 3D Modelling
Kyriacos Themistocleous ✉
( )
Cyprus University of Technology, 2-6 Saripolou, 3036 Limassol, Cyprus

[email protected]
Abstract. Social media can be used as a new source of information by archae‐

ologists and cultural heritage experts to access cultural heritage-related videos for
creating 3D models using Structure for Motion techniques. There is a vast amount
of data now available on social media, which are posted every day on the internet.
However, there is confusion regarding if such data is considered fair use, public
domain, creative commons or copyrighted. Indeed, social medias, such as Face‐
book, Twitter, Instagram and YouTube, have different regulations regarding
ownership and republishing videos. This paper will discuss how social media can
be used for cultural heritage research, especially if video data is used, transformed
or repurposed for 3D modelling.
Keywords: Social media · Video · Documentation · Cultural heritage · Fair use ·

Copyright · Public domain · Creative commons
1 Introduction
There is a wealth of information on social media, including images and videos of cultural
heritage sites, which provides a unique opportunity to cultural heritage specialists. From
photos uploaded to Facebook by tourists to videos taken with UAVs posted to YouTube,
cultural heritage specialists have a plethora of data that can be used study cultural
heritage sites around the world. However, the issue of copyright remains a constant
question. With the use of state of the art photogrammetry techniques, video can be used
to create 3D models [1, 2]. This paper will examine the implications of copyright and
fair use rules for cultural heritage specialists regarding transforming videos posted from
social media into 3D visualization models.
2 Social Media
In recent years, social media is proving to be a very valuable resource for researchers
from a variety of disciplines [3]. The photos and videos available through social media
constitute a valuable source of information for Cultural Heritage professionals. Indeed,
there are several projects that incorporate open data to promote the digital preservation
of lost cultural heritage due to natural disasters, wars, military campaigns, and terrorism
using social media data in a cooperative, open-source project, including Project Rekrei

DOI: 10.1007/978-3-319-48496-9_40
514 K. Themistocleous
(http://projectmosul.org), the Zameni Project (http://zamaniproject.org), the Democra‐

tization of Sciences project (http://aist.usf.edu) and the digital reconstruction of the
Bamiyan Buddhas [4]. It is now possible to use video imaging from social media such
as YouTube, Facebook and Twitter to create 3D models of cultural heritage architectures
and monuments [2, 5]. Indeed, the 3D-ICONS project (3dicons-project. eu/eng/) digi‐
tises in 3D architectural and archaeological monuments and buildings identified by
UNESCO as being of outstanding cultural importance.
Social media are online communication channels that allow people, companies and
other organizations to create, share, or exchange information, ideas, images and videos
in virtual communities and networks. Social media has transformed our means of
communication by providing instant information to publish and publicize almost
anything [6]. Digital technology and the new ways to use and reuse content are chal‐
lenging society’s notion of what is fair use and what is copyrighted [7]. Social
networking sites such as Facebook, Twitter, YouTube, etc. allow users to share their
writings, photos, videos and other creations with the world. Using open media such as
YouTube, Facebook, Tumblr, Spotify, Pinterest, or personal blogs of various types,
billions of online users are sharing, linking, embedding, cutting and pasting as common
practice. Today, every social media user is a publisher of sorts and many publish without
consideration of existing copyright laws.
One of the difficulties in using social media for cultural heritage purposes concerns
the issue of copyright and fair use. There is a vast amount of data now available on social
media, which are posted every day on the internet. As well, sharing text on social media
is an accepted part of social media and to do so would not be an infringement of copyright
unless the text comes from a third party source. However, there is confusion regarding
copyright and the risk of infringement with social media data and if such data is consid‐
ered fair use, public domain, creative commons or copyrighted. Indeed, social media
such as Facebook, Twitter, Instagram and YouTube have different regulations regarding
ownership and republishing images and videos. In order to understand the importance
of copyright in social media, it is important to first define what is copyright, fair use,
public domain and creative common and how this applies to using videos from social
media to create 3D models for cultural heritage purposes. As well, it is important to
realize that the laws governing copyright, fair use, public domain and creative common
differ depending on the country. The majority of laws referring to copyright come from
the United States, the United Kingdom and Australia. Due to the different legal regula‐
tions regarding copyright, fair use, etc. between different countries as well as the different
social media terms of use, cultural heritage specialists should verify with their local laws
prior to publication of their work.
2.1 Copyright
Copyright refers to the protection given to any created work from being copied or
distributed without permission. Copyright attaches as soon as the original work is
created, and applies to both published and unpublished works. Copyright is an automatic
right and does not require the author to file special paperwork, as is the case for trademark
and patent. Derivative works, where an original created work is modified through
Debate and Considerations on Using Videos 515
additions and changes (such as adapting a book for film, or translating a document) are
also covered by copyright.
Copyright covers works that are created or shared on social media websites.
However, the website’s terms and conditions may change the author’s rights to the work.
For example, Facebook, YouTube, Google, Twitter, Flickr and Tumblr gives itself the
right to use anything that is posted, including photos, videos, art work and text.
According to YouTube, copyright owners of the posting can use Content ID to identify
and manage their content on YouTube to protect against copyright infringement.
However, YouTube offers Fair Use protection for posting that borrow small bits of
material from an original work, are transformative in nature and do not affect the market
of the original product. In this case, copyright is not violated since the video is not
reproduced by any means but instead select content is transformed into another media.
2.2 Fair Use

Fair Use refers to the legal right to use copyrighted work in a reasonable manner without
the owner’s consent for purposes such as teaching, scholarship and research. The
purpose of the Fair Use Doctrine is to allow for limited and reasonable uses as long as
the use does not interfere with owners’ rights or impede their right to do with the work
as they wish [8]. Fair use is not the same as free use. Fair use is a legal exception to the
exclusive rights an owner has for his or her copy-righted work.
There are four factors that must be considered when making a fair use claim:
• Purpose - The purpose of the use
• Nature - The nature of the copyrighted work
• Amount - The amount and substantiality of the work that is used in relation to the
work as a whole
• Effect - The effect on the potential market for or value of the copyrighted work
Fair use applies if the original image is repurposed or transformed in such a way that
there is a complete change in meaning and purpose. A particular use of a copyrighted
work may be considered transformative if it uses the copyrighted work in a completely
new way or with a new purpose that the original work’s creator never intended. When
the purpose for using the copyrighted content is not commercial and where the use is
somehow transformative from the copyright holder’s original use, a fair use argument
is strong. The purpose for linking on social media is usually non-commercial and is
transformative, in that the purpose of sharing interesting content is different than the
copyright holder’s purpose for creating and placing it online. Additionally, because the
use is transformative, the act of sharing links to copyrighted content likely will not have
a negative effect on the copyright holder’s market [9]. For example, using videos of a
cultural heritage site to create a 3D model can be considered transformative use. The
most important factor in fair use is purpose- is the use transformative from the original
use? Courts are more likely to uphold the fair use doctrine if the creation is transforma‐
tive, meaning that the end product adds something new, with a different character,
expression, meaning or message.
Figure 1 provides a guideline for Fair Use for using images to create 3D models.
Fig. 1. Guidelines for Fair Use for using images to create 3D models
According to the fair use criteria, using social media for research or educational use,
as is often used by cultural heritage specialists, is acceptable and does not constitute
copyright infringement. Fair use applies in cases where a portion of the images are used
and transformed in such a way that they no longer resemble the images, such as trans‐
forming a video into a 3D model. Although fair use applies to such cases, it is not
mandatory but it is considered good practice to reference the source of the image or
video (if known).
2.3 Creative Commons
Creative commons refer to material that are copyrighted but the creator has put provision
on their use. For example, a creative commons license might stipulate that the image
can be used as long as it is not changed in any way. As of 2015, there were 1 billion
creative commons works on the internet free for the public to reuse [10]. For example,
in the case of YouTube, users can check the video license as “Creative Commons Attri‐
bution License,” which means users can download and reuse the video legally.
2.4 Public Domain

Public domain refers to material that no long has copyright restrictions and no one owns
the copyright. It refers to ideas, facts, and works that have not been fixed in a tangible
medium, works created by the government are not eligible for copyright protection
Works in the public domain are free to be copied and used without restriction. When a
work’s copyright expires, it also becomes a part of the public domain. Copyright does
not apply to works in the public domain; words, names, slogans or short phrases (those
may have protection in trademark law); blank forms; works that are not original; and
government works. Therefore, if the work is not protected by copyright, then there is no
concern whether the Fair Use Doctrine will apply [8].
Figure 2 provides a detailed flowchart for how to decide to use images from another
source for 3D modeling.
Fig. 2. Flowchart for deciding to use images for 3D modeling

3 Methodology
During the last 20 years, cultural heritage and archaeology have experienced a technical
revolution [11] as researchers use the combination of aerial imagery for 3D reconstruction
of the cultural heritage sites. Advances in 3D scanning techniques, drone use and tourist’s
online postings and videos are giving cultural heritage experts and archaeologists a new set
of tools to prevent the permanent loss of cultural artefacts and landscapes. Today, most of
the videos uploaded onto social media are high resolution, which facilitates the use of this
method. These techniques provide a set of new tools for cultural heritage experts to capture,
store, process, share, visualize and annotate 3D models in the field [12–19]. The implemen‐
tation of image-based 3D modeling provides a time- and cost-effective manner for digital
documentation [20–23]. Advances in the fields of Photogrammetry and Computer Vision
have led to significant breakthroughs such as the Structure from Motion (SfM) algorithm
which creates 3D models of objects using sequences of 2D images [1, 2]. There is also a
wealth of research available for using uncalibrated image sequences from video and without
camera calibration to create 3D models [2].
Figure 3 shows an example of the methodologies to create a quality geo-referenced
digital surface model (DSM) using videos of a cultural heritage site from social media
using the SfM algorithm [2]. To generate 3D models from social media, a sequence of
images need to be captured from a video [2]. The captured images then undergo quality
analysis and distortion correction, in order to be aligned and related to each other to get
a number of corresponding points to determine the geometric relationship between the
images. A 3D point cloud is then generated and modelling and texture mapping are
applied. The model can be geo-referenced using Ground Control Points to display the
completed 3D model according to scale. Due to the lack of field measurements, the 3D
models can only be used for visualization purposes and not for accurate documentation.
Fig. 3. Methodology for creating 3D models from social media images [2]
In the 3D reconstruction, the image undergoes several modifications and the final
product is completely transformed from the original. As mentioned above, the process
of using the SfM method changes the video, which is the original media source, and
completely transforms and changes the media, so it clearly falls within into the fair use
guidelines. According to the fair use guidelines (Figs. 1 and 2), the transformation from
video to 3D model described constitutes fair use, since the images captured from video
are repurposed and transformed by digitally changing the distortion, balance, tone,
exposure, contract and clarity to create a new product and media, which is the 3D model.
Also, the sequence of images captured and used to create the 3D model are from a small
portion of the video and are digitally transformed in order to change the media to a 3D
model, which conforms to the fair use guidelines in Figs. 1 and 2. The manipulation of
the images captured from the original video produces a product that is significantly
different from the original media; therefore, it can be regarded as a new work.
4 Conclusions
Although social media provides an opportunity for cultural heritage experts to have
access to a vast amount of data available on the internet, the issue of copyright remains
a constant question regarding how to use this information. This paper clarifies the
difference between copyright, fair use, public domain and creative commons for cultural
heritage experts who are using videos posted to social media to create 3D models. The
use of videos from social media to create 3D models conforms to the fair use guidelines,
since the sequence of images used are transformed to a new media and purpose. The
SfM technique can be used so that videos of cultural heritage sites from social media
can be transformed into 3D models by adhering to fair use guidelines, since the videos
undergoes several modifications and the final product is completely transformed from
the original. The final result, which is a 3D model, can be considered a new work since
it undergoes a specialized process and is transformed from the original media.
References
of tangible cultural heritage objects from web-retrieved images. Int. J. Herit. Dig. Era. 3(2),
431–452 (2014)
2. Themistocleous, K.: Model reconstruction for 3D vizualization of cultural heritage sites using
open data from social media: the case study of soli cyprus. J. Arch. Sc. (2016). doi:10.1016/
j.jasrep.2016.08.045
3. Kietzmann, J., Silvestre, B., McCarthy, I., Pitt, L.: Unpacking the social media phenomenon:
towards a research agenda. J. Pub. Aff. 12(2), 109–119 (2012)
4. Grun, A., Remondino, F., Zhang, L.: Photogrammetric reconstruction of the great Buddha of
Bamiyan, Afghanistan. Photogr. Rec. 19(107), 177–199 (2004)
5. Alsadik, B., Gerke, M. Vosselman, G.: Efficient Use of video for 3D modelling of cultural
heritage objects. BISPRS Ann. Photogramm. Remote Sens. Spatial Inform. Sci., vol. II-3/W4,
2015 PIA15 + HRIGI15 (2015)
6. Cuddy, R.H.: Copyright issues for social media (2013). https://www.legalzoom.com/articles/
copyright-issues-for-social-media
7. Rodriguez, J.E.: Social media use in higher education: key ares to consider for educators.
Merlot J. Online Learn. Teach. 7(4) (2011). http://jolt.merlot.org/vol7no4/rodriguez_
1211.htm
8. Hawkins, S.: Copyright fair use and how it works for online images, 3 November, 2011
(2011). http://www.socialmediaexaminer.com/copyright-fair-use-and-how-it-works-for-
online-images/
9. Gutierrez Alm, J.: Sharing copyrights: the copyright implications of user content in social
media Hamline University’s School of Law’s. J. Publ. Law Policy 35(1), 2 (2014).
Transcending Intellectual Property Rights: an Exploration into the Unchartered Territories of
the Intangible, Infringed, and the Internet
10. Creative Commons (2015). https://stateof.creativecommons.org/report/

11. Macheridis, S.: Image-based 3D modeling as a documentation method for Zooarchaeological
remains in waste-related contexts. Ethnobiol. L. 6(2), 242–248 (2015). doi:10.14237/ebl.
6.2.2015.342
12. Ioannides, M., Hadjiprocopis, A., Doulamis, N., Doulamis, A., Protopapadakis, Eft.,
Makantasis, K., Santos, P., Fellner, D., Stork, A., Balet, O, Julien, M., Weinlinger, G.,
Johnson, P.S., Klein, M., Fritsch, D.: Online 4D reconstruction using multi-images available
under open access. ISPRS Ann. Photogramm. Remote Sens. Spatial Inform. Sci., II-5/W1,
169–174 (2013)
13. Themistocleous, K., Agapiou, A., King, H.M., King, N., Hadjimitsis, D.G.: More than a flight:
the extensive contributions of UAV flights to archaeological research–the case study of curium
site in cyprus. In: Ioannides, M., Magnenat-Thalmann, N., Fink, E., Žarnić, R., Yen, A.-Y.,
Quak, E. (eds.) EuroMed 2014. LNCS, vol. 8740, pp. 396–409. Springer, Heidelberg (2014).
doi:10.1007/978-3-319-13695-0_38
14. Themistocleous, K., Agapiou A., Cuca, B. Hadjimitsis, D.G.: Unmanned aerial systems and
spectroscopy for remote sensing applications in archaeology. In: Proceedings of the 36th
International Symposium on Remote Sensing of Environment (ISRSE–36) (2015a)
15. Themistocleous K., Agapiou A., Lysandrou, V., Hadjimitsis, D.G.: The use of UAVs for
remote sensing applications: case studies in cyprus. In: Proceedings of SPIE 9644, Earth
Resources and Environmental Remote Sensing/GIS Applications VI, 96440T (20 October
2015). doi:10.1117/12.2195027 (2015b)
16. Themistocleous, K., Ioannides, M., Agapiou, A., Hadjimitsis, D.G.: The methodology of
documenting cultural heritage sites using photogrammetry, UAV and 3D printing techniques:
the case study of Asinou Church in Cyprus. In: Proceedings of the Third International
Conference on Remote Sensing and Geoinformation of Environment, SPIE 953510 (June 19,
2015). doi:10.1117/12.2195626 (2015c)
17. Themistocleous K., Ioannides M., Agapiou A., Hadjimitsis D.G.: A new approach for
documenting architectural cultural heritage: the case study of Asinou Church in Cyprus. In:
Proceedings of International Conference on Sustainability in Architecture and Cultural
Heritage (2015d)
18. Fiorillo, F., Jimenez, B., Remondino, F., Barba, S.: 3D surveying and modeling of the
archaeological area of Paestum, Italy. Virt. Archaeo. Rev. 4, 55–60 (2012)
19. Eisenbeiss, H.: UAV Photogrammetry, Ph.D. Thesis. Institute fur Geodesie und Photogram-
metrie, ETH-Zurch. Zurich, Switzerland (2009)
20. de Reu, J., de Smedt, P., Herremans, D., van Meirvenne, M., Laloo, P., De Clercq, W.: On
introducing an image-based 3D reconstruction method in Archaeological excavation practice.
J. Arch. Sc. 41, 251–262 (2014). doi:10.1016/j.jas.2013.08.020
21. de Reu, J., Plets, G., Verhoeven, G., de Smedt, P., Bats, M., Cherrette, B., de Maeyer, W.,
Deconynck, J., Herremans, D., Laloo, P., van Meirvenne, M., de Clercq, W.: Towards a three-
dimensional cost-effective registration of the archaeological heritage. J. Arch. Sc. 40, 1108–
1121 (2013). doi:10.1016/j.jas.2012.08.040
22. Dellepiane, M., Dell’Unto, N., Callieri, M., Lindgren, S., Scopigno, R.: Archeological
excavation monitoring using dense stereo matching techniques. J. Cult. Her. 14, 201–210
(2013). doi:10.1016/j.culher.2012.01.011
23. Verhoeven, G., Doneus, M., Briese, C., Vermeulen, F.: Mapping by matching: a computer
vision-based approach to fast and accurate georeferencing of archaeological aerial
photographs. J. Arch. Sc. 39, 2060–2070 (2012). doi:10.1016/j.jas.2012.02.022
Investigations of Low-Cost Systems for 3D Reconstruction
of Small Objects
Thomas P. Kersten ✉ , Daniel Omelanowsky, and Maren Lindstaedt

( )
Photogrammetry and Laser Scanning Lab, HafenCity University Hamburg,

Überseeallee 16, 20457 Hamburg, Germany
{Thomas.Kersten,Maren.Lindstaedt}@hcu-hamburg.de,
[email protected]
Abstract. In this paper geometric investigations are presented, which demon‐

strate the potential of the low-cost recording systems DAVID SLS-1 and Micro‐
soft® Kinect for sustainable use in applications for architecture, cultural heritage
and archaeology. From the data recorded with DAVID SLS-1 and Microsoft®
Kinect 3D models were produced by different programs and these were examined
in relation to handling, quality and reliability in further post processing. For the
investigations a number of 3D objects with different surface forms, including a
test body, were scanned using the structured light system ATOS I 2M from GOM
as references. To compare the results of the Kinect and the SLS-1, digital surface
models of this test body were automatically generated using image-based low-
cost recording systems (Nikon D7000). As a result of these 3D comparisons to
the ATOS reference data a standard deviation of 1.5 and/or 1.6 mm was obtained
with the structured light system SLS-1 and/or with the Kinect, while with the
image-based 3D reconstruction methods of VisualSFM/CMVS a higher standard
deviation of up to 0.2 mm was achieved. Although the introduced low-cost struc‐
tured light system David SLS-1 could not show the geometrical accuracy of a
high end system (ATOS I) of approx. 0.04 mm, it is useful for the 3D recording
of smaller objects (size up to 50 cm) with a reduced accuracy for several different
applications.
Keywords: 3D · Automation · Comparison · Image matching · Modelling · Point

cloud · Reconstruction
1 Introduction
For 3D shape recording of small and complex objects modern systems typically work
according to the triangulation principle. For a long time these systems were only avail‐
able at the expensive, high-end of the market, but in recent years affordable options have
become increasingly available for users with smaller budgets. Current low-cost systems
are available on the market but neither the quality nor the quantity of the recorded data
is often considered in public discussion. However, contributions from Hieronymus et al.
[1], Wujanz et al. [2], Khoshelham [3] and Boehm [4] present test results of 3D sensors
from the low-cost field and the gamer market. Since the market for 3D consumer sensors

DOI: 10.1007/978-3-319-48496-9_41
522 T.P. Kersten et al.
is constantly growing, an examination of this equipment regarding reliability and accu‐

racy for measuring tasks is an obvious requirement. Results of empirical accuracy tests
are presented in [5], in which models from image-based low-cost 3D reconstruction
methods are compared with reference data of higher accuracy.
The applications of so-called low-cost systems are to be found predominantly within
the field of archaeology and cultural heritage, in which the structured light projection
procedure is often used by preference. Thus, Sablatnig and Menard [6], Akca et al. [7],
McPherron et al. [8], Bathow and Breuckmann [9] described that structured light scan‐
ners are widely adopted for these applications, since the contactless procedure is partic‐
ularly well suited for the 3D documentation of small objects. First test results with
Microsoft Kinect were presented by Wujanz et al. [2], Khoshelham [3] and Smisek et al.
[10], while Mankoff and Russo [11] have also reported on experiences with the Kinect
in glaciological, bathymetrical and geomorphological applications.
In the present contribution the potential of both low-cost systems DAVID SLS-1 and
Microsoft® Kinect has been evaluated in comparison to image-based reconstruction
procedures using several test objects as practical examples. Reference measurements of
the different test objects were recorded with a structured light system (GOM ATOS
I 2M).
2 Low-Cost Systems Evaluated
In these investigations a total of three low-cost recording systems was evaluated: (a) the
Microsoft® Kinect [12], (b) the DAVID SLS-1 [13], which both work with an active
(projector) and a passive (camera) sensor using the triangulation method, and (c) an
image-based recording system (using different digital SLR cameras).
The Kinect was used in combination with the ReconstructMe software [14] which
is freely-available for non-commercial applications. This sensor, which originates in the
gaming market, offers a resolution of 640 × 480 pixels with a measuring distance of 0.8
to 3.5 m. The Kinect costs approximately €100. The operational and functional principle
is described in [15]. The SLS-1 is a low-cost structured light system from DAVID Vision
Systems GmbH in Koblenz, which commercially distributed this product including the
software DAVID Laserscanner as a complete 3D recording solution at a price of approx‐
imately €1700 (Nov. 2012). The integrated camera consists of a monochrome CMOS
image sensor with a fixed focal length of 12 mm and an image resolution of 744 × 480
pixels. The projector used in this system, which projects the structured light, is a
commercial Acer K11. The base distance between camera and projector can be deter‐
mined by system calibration in the software DAVID Laserscanner. Although the camera
is monochrome, it is possible to texture the 3D models with colour by projecting several
colours onto the object with the projector and by measuring the respective reflections
from the object surface with the monochrome camera. DAVID software offers the whole
workflow from object recording through scan registration to the export of the model,
while with ReconstructME only recording and data export are possible. In principle, the
measuring volume is unlimited for the MS Kinect, if certain conditions are considered
as described in [14]. With the DAVID SLS-1 the maximum expanse of the object surface
Investigations of Low-Cost Systems for 3D Reconstruction 523
is 30–500 mm. For image-based reconstruction procedures the digital SLR camera
Nikon D7000 (4,938 × 3,264 pixels) was used, which can be very flexibly used for object
recording. The following automatic generation of 3D point clouds and/or 3D object
models was carried out with the open-source software Bundler [16] and PMVS2 [17],
with the free available non-commercial software VisualSFM [18] as well as with the
Autodesk Web service 123D Catch [19]. These three image-based 3D reconstruction
methods are already briefly introduced in [5].
The evaluated recording systems are represented in Fig. 1; for detailed technical
specifications the relevant manufacturers’ web pages are referenced.
Fig. 1. Evaluated low-cost systems (f.l.t.r.): Nikon D7000, DAVID SLS-1 and MS Kinect
3 Evaluation Criteria and Methods
3.1 System Stability

The measuring systems DAVID SLS-1 and MS Kinect can be calibrated and prepared
for data recording (the digitization procedure) within a few minutes by an experienced
user. In order to be able to meet a statement about the stability of the SLS-1, as the first
investigation camera calibration was conducted twenty times every ten minutes. The
interior (intrinsic) and exterior (extrinsic) parameters of the camera are summarized in
the calibration report. The SLS-1 uses a camera model of Tsai [20]. Which includes the
following interior orientation parameters: focal length f, radial distortion coefficient κ1,
scale factor sx and the coordinates of the principle point cx and cy.
Figure 2 illustrates the SLS-1 calibration setup (left) and the computed focal length
C in pixels (px) for the observation period of 200 min, in which 20 calibrations were
carried out with the software DAVID Laserscanner (version 3.4.0).
For system calibration the camera acquires images of an orthogonally constructed
calibration field provided with targets (Fig. 2 left) and whose geometry is known. The
lowest and highest value of the calibrated focal length in the calibration series has a wide
margin of 3.5 pixels, whereby the largest deviation was registered in minute 150. Unlike
the other values, the series of measurements in minutes 30–60, 90, 110, 160 and 200
match each other well since their deviations are only approx. one pixel. In principle all
values are evenly distributed, but a descending trend is clearly observable. Additionally,
Fig. 2. DAVID SLS-1 (software version 3.4.0) – Calibration setup (left) and the variation of the
focal length (right) over the time period of 200 min
observation of the remaining intrinsic parameters (parameters of interior orientation)

took place during these tests. It was observed that the location of the principle point in
x-direction varies by only approximately 2.5 pixels. On the other hand, a span of 4 pixels
is shown in the y-direction and it rises continuously starting from the first measurement.
The scale factor varies around the value 0.084; it has the largest deviation of 0.0001
from the average value in minute 70. Although the temperature environment in the
laboratory was constant, the computed parameters of interior orientation appeared to
vary arbitrarily. This demonstrates a minor instability of the SLS-1, which also might
affect the geometrical results of data recording.
In order to evaluate the stability of the Kinect on the basis of system calibration
results, 20 photographs were manually taken from different positions and were stored
in a log file (protocol of results). This procedure was repeated in a series of five meas‐
urements. The intrinsic parameters of the Kinect include the camera constant in x and
y-direction (fx/fy) and the coordinates of the principle point (px/py). The results of the
series of measurements are illustrated in Fig. 3, which cannot be compared with the
series of measurements in Fig. 2.
Fig. 3. MS Kinect – calibration pattern (left) and the variation of the focal length (right)
The span of the focal length is up to 10 pixels (see measurement series 3 and 4 in
Fig. 3) for all series of measurements, whereby the largest deviation is approximately
3.5 pixels between fx and fy. On the other hand the values of the measurement series 1,
2 and 5 vary only approximately 1 pixel. Exactly the same results were achieved for the
coordinates of the principle point. The largest deviation compared to the average value
is registered in measurement 4 at approx. 4 pixels. Due to the fact that the values of
interior orientation vary substantially despite calibrations under constant laboratory

conditions, high system stability can also not be certified for the Kinect.
Investigations with DAVID SLS-1 and MS Kinect have demonstrated that the cali‐
bration results seem very arbitrary and are often inexplicable. However, the following
measurements and 3D comparisons to reference data show that the differences for small
objects are in the sub-millimetre range. Both the camera and the projector of the DAVID
system were calibrated. The workflow and the results of the projector calibration are
described in detail in [21].
3.2 Repeatability (Precision)

Repeatability (precision) is a criterion relating to the quality of a measuring procedure.
It is also called as internal accuracy of a measurement, and is determined by the repetition
of measurements. Using a gypsum figure, whose body size corresponds to the usual
recording volume (91 mm × 156 mm × 91 mm) of the system, the repeatability of the
DAVID Systems was analysed.
In addition, the data capture of the gypsum figure was performed every six minutes
during a time period of two hours. These six minutes correspond to the recording time
of a real scanning object. In this way 19 different models were computed for analysis in
Geomagic, each consisting of approximately 104,000 triangle points. A precision of
0.007 mm resulted from the mean of the average deviations of the 19 difference models.
The largest deviations to the master scan (first scan of the SLS-1 is set as reference) of
+0.032/−0.041 mm occur after half of the recording time and/or measurement series.
In order to guarantee that the master scan meets the precision needs of the SLS-1 (0.2 %
of the measuring volume according to the manufacturer specifications), a 3D comparison
to the reference model of the ATOS I was generated.
Fig. 4. Deviation between master scan (SLS-1) and scan 10 (left) as well as 3D comparison
(centre and right) of both models (green = better than 0.3 mm) (Color figure online)
In Fig. 4 one can see that the average deviations of 0.23 mm meet the precision needs
of the SLS-1. The maximum deviation of 1 mm at the edge results from measuring noise
or light reflection. This can be ignored for the data analysis, since boundary regions can
be smoothed and shrunken by edge operators within DAVID Software to ensure that
they do not distract from the finished 3D model. In Fig. 4 (right) the 3D comparison
makes it obvious that the generated model shows patch errors (registration of the scans)
on the back of the gypsum figure with particularly bad results in the shoulder region
where the deviations are up to a millimetre.
3.3 Analysis of Deviation by Comparison with Reference

In order to give a statement about the geometrical quality of the three systems used, the
test body “Testy”, which was introduced by Reulke and Misgaiski [22], from the Institute
for Computer Science of the Humboldt university in Berlin was used (Fig. 5). First, as
a reference, the test body was recorded in 52 scans with the ATOS I (17 mm lens)
(approx. 225,000 points). In particular, the surfaces of the model in the indentation and
in the twist were difficult to scan.
Fig. 5. Test body testy from the Humboldt University in Berlin (left and right) and the reference
scan using ATOS I 2M (centre)
Subsequently, the test body was measured with the SLS-1 by twelve scans, which
were registered to each other using the David software. During measurement the same
problems occurred as with the ATOS I due to the occlusion of certain model areas. The
scanning was carried out in approx. 15 min, while the production of a 3D model which
followed took 105 min. For the recording with the system Kinect/ReconstructMe, seven
attempts on a high performance computer (Intel Xeon CPU E5540 with 2.53 GHz, 12 GB
RAM, NVIDIA GeForce GTX 690) were necessary in order to reconstruct the test body
in real time. The automatic registration of the scans often failed during recording
meaning that the recording procedure had to be aborted. However, with an error free
recording process the 3D model of Testy was generated within three minutes. Co-regis‐
tration as a post-processing step is not possible since the two data records cannot be
registered on-line. The acquisition of 54 photos with the Nikon D7000 (18 mm lens) for
the image-based 3D reconstruction procedure with Bundler/PMVS2, VisualSFM and
123D Catch was conducted within 15 min. For scaling of the object, two points on the
test body were measured with a total station Leica TCRP 1201+. The distance between
the two object points was determined with a standard deviation of 0.2 mm. For the
subsequent scaling of the different models Geomagic Qualify was applied using the
computed distance between the two points. The speed of the subsequent 3D reconstruc‐
tion of Testy depended upon soft- and hardware. A standard computer (Intel core 2 duo
CPU T7700 with 2.40 GHz, 12 GB RAM and a NVIDIA GeForce 8600M GT) needed
approx. 60 min with VisualSFM for generation of the 3D point cloud, while with
Bundler/PMVS2 a point cloud was produced after 480 min. On the other hand after the
photos were uploaded on the server the Autodesk web service 123D Catch needed
approx. 10 min to make the computed 3D model available for download. In order to
support the measuring algorithms for image orientation by providing appropriate
textures in the object background, the photographs were taken after putting the test body
on a newspaper (Fig. 5). In post processing triangle meshes were computed in Geomagic
using the point clouds from Bundler/PMVS2 and VisualSFM. Finally, all meshed
models of Testy were compared with the reference data from ATOS I in Geomagic
Qualify. Firstly a rough registration of the five models to the reference model was
manually undertaken using identical points. This was later refined using the ICP algo‐
rithm with the best fit method. The results of 3D comparisons between the reference
model from the high-end system ATOS I, the two evaluated low-cost systems and the
image-based reconstruction procedures are represented in Fig. 6.
Fig. 6. Generated 3D models of testy in comparison to ATOS I 2M (f.l.t.r.): DAVID SLS-1, MS

kinect, VisualSFM/CMVS, Bundler/PMVS2 and autodesk 123D catch (green < 0.8 mm) (Color
figure online)
In the results of 3D comparison between SLS-1 and the reference systematic effects
(negative deviations) are clearly consistently distributed over the entire body (see
Fig. 6 left). It is to be assumed that these systematic deviations are caused by a scaling
error in the SLS-1 data. In contrast the deviations between the Kinect model and the
reference are unevenly distributed, since the data capture with ReconstructMe was also
carried out very unevenly (recording start and stop). However, the models from image-
based 3D reconstruction procedures show only very small deviations.
In Table 1 the results of the 3D comparison, which were generated in Geomagic, are
summarised. They show that the image-based reconstruction procedures achieved
significantly better results for maximum and average deviation as well as for standard
deviation as the two tested systems SLS-1 and MS Kinect. Similar good results for
image-based reconstruction procedures were shown by Kersten and Lindstaedt [23] for
small to medium sized objects in archaeology and cultural heritage.
Table 1. Deviations to reference for the models of the test body Testy generated by different
systems [mm].
System/Software # triangles Max. dev. Av. dev. + Av. dev. − Std. dev.
DAVID SLS-1 1.650.404 21.5 2.3 1.5 1.5
Kinect/ReconstMe 389.628 15.3 0.9 0.9 1.6
VisualSFM 464.246 2.7 0.1 0.1 0.2
Bundler/PMVS2 405.980 4.0 0.1 0.1 0.3
123D Catch 14.034 5.2 0.5 0.3 0.7
Fig. 7. Generated 3D models with DAVID SLS-1 in comparison to ATOS I 2M – wheel hub
(left, green = ±0.5 mm), penguin (centre, green = ±0.2 mm) and Einstein bust (right,
green = ±0.3 mm) – to show systematic deviations (Color figure online)
Additionally, three more objects were used for a 3D comparison of the results from
the DAVID SLS-1 and the data from the ATOS I 2M (reference). The three objects were
composed of different materials: cast iron for the wheel hub, bronze for the penguin,
and gypsum for the Einstein bust. These objects were scanned as follows: wheel hub –
70 scans with ATOS, 24 scans with SLS-1, penguin – 70 scans with ATOS, 6 scans with
SLS-1, and Einstein bust – 21 scans with ATOS, 17 scans with SLS-1. The results of
the 3D comparison are illustrated in Fig. 7 and summarised in Table 2. The difference
plots in Fig. 7 show that the problematic surface areas for scanning with the SLS-1 are
the edges of the object. Nevertheless, some systematic effects (indicated in blue) are
also illustrated in the difference plots.
Table 2. Deviations to reference for the models of three test objects generated by DAVID SLS-1
[mm].
Test body Size [cm3] # triangles Max. dev. Av. dev. + Av. dev. − Std. dev.
Wheel 23 × 12 × 23 4.057.057 11.1 0.5 −0.7 1.2
Penguin 22 × 46 × 28 2.229.541 16.5 0.2 −0.2 0.4
Einstein 11 × 16 × 10 912.320 −13.6 0.3 −0.4 0.6
3.4 Polygon Decimation

Usually the modelling of object surfaces as polygon networks creates large data
volumes, which makes representation of such data on terminal devices with low
performance (e.g. Smartphones with 600 MHz processor) or on the Internet problematic.
For this reason polygon decimation of meshed 3D models is the solution for such a task
to guarantee fast access to the data. Results on polygon decimation of meshed models
from terrestrial laser scanning data were already published in [24]. These results showed
that one can reduce the data set of 3D models up to 10 % without having significant
losses to geometrical or visual quality.
To verify this, the test body Testy was examined in relation to polygon decimation
using the data from the DAVID SLS-1. A file size of approx. 120 MB (1.7 million
triangles) corresponds to 100 % of the original data recorded for Testy. The percentage
of polygon decimation can be defined by the user in the DAVID Laserscanner software.
Up to a polygon decimation of 20 % (approx. 290,000 triangles) there is no significant
geometrical deviation to the original. However, the visual comparison of the test body
reduced to 20 % already shows smoothing effects at edges and at the targets. Neverthe‐
less, the meshed models can be reduced up to 20 % of the volume of data without accu‐
racy losses. The results of the polygon decimation are represented in Fig. 8 for four 3D
models of the test body Testy (100 %, 20 %, 6 % and 4 %).
Fig. 8. Results of polygon decimation (f.l.t.r.): Testy 100 %, 20 %, 6 % and 4 %.
4 Conclusion and Outlook
In this paper geometrical investigations under laboratory conditions showed that the two
low-cost systems DAVID SLS-1 and Microsoft Kinect generate significantly worse
results compared to image-based reconstruction procedures using digital SLR cameras.
The 3D models generated from image-based reconstruction procedures are characterized
by a high degree of automation and by very good quality. However, the two evaluated
low-cost systems offer the advantage of on-line control of completed recording of the
object and/or of the object space during the digitization process. On the other hand
image-based reconstruction procedures automatically compute 3D point clouds and/or
3D surface models without direct access by the user meaning that gross errors become
visible only after the computation from the photos. Nevertheless, the quality of the SLS-1
results corresponds to the accuracy after system calibration specified by the manufac‐
turer (0.2 % of the measuring volume). Unfortunately, no information about the algo‐
rithms used in the DAVID Laserscanner software is available to the user (e.g. registra‐
tion, triangle meshing, etc.). The complete package DAVID SLS-1 is suitable for users
who would like to produce digital 3D models with limited accuracy of small objects for
various applications and within a short time. A further decisive criterion for the purchase
of a SLS-1 is the attractive price, which is 25 times less than for high-end products (e.g.
ATOS I 2M). The results of the system Microsoft Kinect/ReconstructMe were better
than from the SLS-1; however there is still substantial optimization potential in hard-
and software for both systems. Efficient and successful 3D object recording with the
Kinect requires repetition and/or practical experience by the user. Thus, several
approaches are often necessary in order to achieve a useful result. Furthermore, a good
computer with high performance graphics processor is necessary when using Kinect/
ReconstructMe in order to obtain results at all. Generally, this low-cost system offer an
economically accessible product for many users for providing contactless 3D object
recording for digitization of objects (including in the field of education).
References
1. Hieronymus, J., Misgaiski, M., Reulke, R.: Genauigkeitsvergleich von 3D-Sensoren aus dem
Freizeit- und Spielemarkt. Photogrammetrie – Laserscanning – Optische 3D-Mess-technik,
Beiträge der Oldenburger 3D-Tage 2011, T. Luhmann/C. Müller (eds.), Wichmann Verlag,
Heidelberg, pp. 232–234 (2011)
2. Wujanz, D., Weisbrich, S., Neitzel, F.: 3D-Mapping mit dem Microsoft® Kinect Sensor –
erste Untersuchungsergeb-nisse. Photogrammetrie – Laserscanning – Optische 3D-Mess-
technik, Beiträge der Oldenburger 3D-Tage 2011, T. Luhmann/C. Müller (eds.), Wichmann
Verlag, Heidelberg, pp. 274–283 (2011)
3. Khoshelham, K.: Accuracy analysis of kinect depth data. In: The International Archives of
the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XXXVIII-5/
W12, pp. 133–138 (2011)
4. Böhm, J.: Accuracy investigation for structured-light based consumer 3D sensors.
Photogrammetrie – Fernerkundung – Geoinformation, (2), 117–127 (2014)
5. Kersten, T.P., Lindstaedt, M.: Automatic 3D object reconstruction from multiple images
for architectural, cultural heritage and archaeological applications using open-source
software and web services. Photogrammetrie – Fernerkundung – Geoinformation, (6), 727–
740 (2012a)
6. Sablatnig, R., Menard, C.: Stereo and structured light as acquisition methods in the field of
archaeology. In: Fuchs, S., Hoffmannm, R. (eds.) Mustererkennung 1992, pp. 398–404.
7. Akca, D., Remondino, F., Novák, D., Hanusch, T., Schrotter, G., Gruen, A.: Recording and
modeling of cultural heritage objects with coded structured light projection systems. In: 2nd
International Conference on Remote Sensing in Archaeology, Rome, Italy, December 4-7,
pp. 375–382 (2006)
8. McPherron, S.P., Gernat, T., Hublin, J.J.: Structured light scanning for high-resolution
documentation of in situ archaeological finds. J. Archaeol. Sci. 36(1), 19–24 (2009)
9. Bathow, C., Breuckmann, B.: High-definition 3D acquisition of archaeological objects: an
overview of various challenging projects all over the world. In: 23rd CIPA Symposium, pp.
12–16 (2011)
10. Smisek, J., Jancosek, M., Pajdla, T.: 3D with kinect. In: Fossati, A., Gall, J., Grabner, H., Ren,
X., Konolige, K. (eds.) Consumer Depth Cameras for Computer Vision, pp. 3–25. Springer,
London (2013)
11. Mankoff, K.D., Russo, T.A.: The Kinect: a low-cost, high-resolution, short-range 3D camera.
Earth Surf. Proc. Land. 38(9), 926–936 (2013)
12. Microsoft: Kinect for Windows® (2015). www.microsoft.com. Accessed 20 Sep 2015
13. David: DAVID 3D Scanner (2012). www.david-laserscanner.com. Accessed 6 Dec 2012
14. Profactor: technology to capture 3D models in real time (2016). www.reconstructme.net.

Accessed 21 Mar 2016
15. Garcia, J., Zalevsky, Z.: Range Mapping using speckle decorrelation. United States Patent,
Patent No. US 7,433,024 B2 (2008)
16. Snavely, N., Seitz, S.M., Szeliski, R.: Modeling the World from internet photo collection. Int.
J. Comput. Vis. 80(2), 189–210 (2008)
17. Furukawa, Y., Ponce, J.: Accurate, dense and robust multi-view stereopsis. IEEE Trans.
Pattern Anal. Mach. Intell. 32(8), 1362–1376 (2010)
18. Wu, C.: VisualSFM: A Visual Structure from Motion System (2011). http://ccwu.me/vsfm/
doc.html#dep. Accessed 21 Mar 2016
19. Courchay, J., Pons, J.P., Monasse, P., Keriven, R.: Dense and accurate spatio-temporal multi-
view stereovision. In: Zha, H., Taniguchi, R., Maybank, S. (eds.) ACCV 2009. LNCS, vol.
20. Tsai, R.Y.: An efficient and accurate camera calibration technique for 3-D machine vision.
In: IEEE International Conference on Computer Vision and Pattern Recognition, Miami,
USA, pp. 364–374 (1986)
21. Omelanowsky, D.: Untersuchungen der Low-Cost 3D-Aufnahmesysteme DAVID SLS-1 und
Microsoft Kinect. Unpublished Master Thesis in Master Study Programme Geomatics,
HafenCity University Hamburg, January (2013)
22. Reulke, R., Misgaiski, M.: Test body “Testy” for laser scanning and optical systems.
Photogrammetrie – Fernerkundung – Geoinformation, (6), zum Titelbild (2012)
23. Kersten, T.P., Lindstaedt, M.: Image-based low-cost systems for automatic 3D recording and
modelling of archaeological finds and objects. In: Ioannides, M., Fritsch, D., Leissner, J.,
Davies, R., Remondino, F., Caffo, R. (eds.) EuroMed 2012. LNCS, vol. 7616, pp. 1–10.
24. Kersten, T.P.: 3D scanning and modelling of the bismarck monument by terrestrial laser
scanning for integration into a 3D city model of hamburg. In: Ioannides, M., Fellner, D.,
Georgopoulos, A., Hadjimitsis, D.G. (eds.) EuroMed 2010. LNCS, vol. 6436, pp. 179–192.
Digital Documentation: Villa Borghese
Martina Attenni1, Cristiana Bartolomei2, and Alfonso Ippolito1(&)

1
Sapienza University of Rome, Rome, Italy
{martina.attenni,alfonso.ippolito}@uniroma1.it
2
Alma Mater Studiorum University of Bologna, Bologna, Italy
[email protected]
Abstract. The knowledge of historical and architectural heritage is today

reinforced by the growing use of digital instruments serving the purpose of
documenting and disseminating data. The development techniques to build 3D
models made them pivotal elements in popularizing information on objects on
the scale of architectonic structures. Digital archives supplement 3D models
with heterogeneous data (2D models, images, texts, video materials, biblio-
graphical documents) with the purpose to preserve, evaluate and popularize
cultural heritage (CH) by devising an open system of knowledge. This study
puts forward a critical operative method and some guidelines to record, con-
struct, manage, visualize and navigate 3D models with a view to achieving a full
comprehension of the architecture in their own context, permitting to discover
their inter-relationships through a digital archive.
Keywords: Villa Borghese Cultural heritage Survey 3D/2D models

Digital archives
1 Introduction
Evaluation and preservation of CH is connected with the innovative processes of

gaining, managing and using knowledge. For several years now, data archiving has
been a central issue insofar as works of cultural heritage are concerned. It is precisely in
this domain that the methodologies of digital 3D models representation were signifi-
cantly applied. The ever-growing number of digital models as well as the diversity of
processes and objects involved impose the necessity to define new modalities for
documentation, managing and sharing information. Nowadays the techniques and
instruments of digitalization have made possible the large scale production of 3D
objects belonging to CH. The main feature that transforms physical objects into digital
ones involves the construction of photorealistic 3D models close to geometric, metric,
structural, chromatic models geometrically, metrically, structurally, chromatically
closely resembling real objects. These models provide a much more articulated and
complex representation than others because they can be used as bases to construct
knowledge systems. Digital visualization makes possible an integrated interpretation of
heterogeneous data and a redefinition of the very concept of representation by linking it
to the concept of information. 3D models are the central point as representation of
existing artifacts and as metaphors for navigation inside other types of data. This study

DOI: 10.1007/978-3-319-48496-9_42
534 M. Attenni et al.
describes an integrated system to record, construct, manage, visualize and navigate 3D

reality based models of architectural object into a digital archive. Constructing such a
system has to overcome problems of creating a model based digital archive: how to
obtain 3D reality based models of objects which are different but homogeneous as to
their scale [1], uncertainty level and visualization typology; how to structure the 3D
database able to display documentary materials (documents, photos, drawings) digi-
tally; which criteria to apply in order to make the procedure repeatable in case studies
with analogous characteristics; the standardization of 3D/2D model. Surveying oper-
ations, indispensable when the aim is to construct models of real objects are linked to
the characteristics of the case studies analyzed, on the level of detail representation and
of communicative aims. In order to address these problems, we defined, a-priori, and
verified, a-posteriori, standardized characteristics of 3D and 2D models, able to eval-
uate both the quality and the fidelity with accurate pre-defined metric standards. 3D
models were built with the purpose to identify the architecture and their related
resources (images, 2D models, text, etc.) as elements connected with the 3D geometry.
Through the recent surveying technologies is possible to obtain models that include all
fine details of the original artifact such as a photo-realistic representation of the surface.
In recent years there have developed new research areas concerned with the speedy
acquisition of data that contain a high quantity of information; use of low cost
instruments that ensure results comparable to those obtained with the 3D laser scanner
within a single acquisition and restitution campaign. This is precisely the principal area
of enquiry of our research, which aims primarily at realizing models obtained through a
survey that integrates traditional and low cost methodologies for non-contact survey-
ing. The study seeks to integrate the process of surveying with that of cataloguing by
structuring out models obtained through digital photographic images in a documentary
archive.
2 Villa Borghese Architectures: Data Capture and Processing
The Villa Borghese covers the area of approximately 80 ha in the very center of Rome.
Its construction as a suburban residence started around the year 1606 but today it is an
open-air museum containing buildings, architectural complexes, monuments and
fountains, all works of outstanding baroque, neoclassical and eclectic artists. The study
of the architecture of Villa Borghese started with a careful and detailed cataloguing of
existing structures. The park is dotted with elements representing various architectural
typologies of grand, medium and small scale. The buildings erected for the purpose of
representation function today as venues of cultural events and intellectual activity: art
galleries, museum, academies. Others, conceived as forms of embellishing the gardens,
have not changed their original function and identify important places in the park like
entrances or intersections between main alleys. The present research concerns only
objects of the medium scale (entrances or architectural décor) and on the small scale
(fountains and sculptures). The choice is related to the experimental aspect of the
surveying methodology adopted, considered especially suitable for artifacts of this
type.
Digital Documentation: Villa Borghese 535
The survey operation, as a process involving extensive knowledge1, aims at the

accurate representation of the analyzed subject and its goal goes beyond the consid-
eration of the technique and the specific instrument, making the ultimate result of this
process broader and more amplified. For this reason, drafting a procedure based on the
interaction between direct and indirect2 high cost and low cost methodologies for the
elaboration of 2D and 3D models would be a topic of great interest. The definition and
managing of the 3D model become the crucial point of the problem for the solution of
which it is fundamental to define an operative methodology that could be referred to in
a standardized manner. The concepts of data quality3 and scientificity4 [2] alongside
that of uncertainty control turn out to be fundamental in the processes of acquisition
and elaboration. Nowadays the integration of surveying methods has proved to be
necessary, and can now be considered a standard practice. It is thanks to this integration
that we are capable of understanding the object of study and analysis, in both general
and detailed terms. Surveying is thus to be intended as a rigorous methodological
process that, through operations of selection, measurement and representation of
important points, is capable of describing the geometric-spatial, dimensional and for-
mal qualities of the object of study and, in absolute, allowing for the achievement of the
profound awareness of the aforementioned object. Preparing a survey campaign for a
huge architectonic site such as the one discussed here obviously entails an initial stage
of rigorous planning and analysis in order to optimize the information that is then
acquired. This stage must be regarded as fundamental for two reasons: first, because it
is precisely the results of this preparatory effort that will provide the material which will
be worked on to produce the survey’s deliverables; second, because this is the only
stage in which the data will have all of the numerical information deriving from the
measurement operations, and that can always be used as a critical data as it has not been
processed or interpreted in any way. This work discusses the possibilities offered by the
integration of heterogeneous methods, traditional and innovative, for massive
1
René Descartes distinguishes between normal knowledge, achieved only by our senses, and profound
knowledge, achieved by scholars using only study methods and techniques that can demonstrate to
the mind what is precluded to the senses.
2
All operations that envisage the gathering of knowledge need to have a reference framework both
vis-à-vis the data acquisition method (surveying) and the selection, processing and restitution of the
acquired data (survey).
3
Quality is here understood as a description of contingent or permanent properties of an object. It
designates any concretely determined formal aspect of a given reality. Thus, a meteorological setting
is qualitatively described by an aggregate of parameters that define the characteristic of measure:
uncertainty, repeatability, accuracy.
4
“The stage of data acquisition […] should always be considered as an application of a scientific
method […] based on a collection of empirically observable and measurable data […] which can be
archived […] subjected to external verification […] to be validated; the applied procedure should be
replicable in order to acquire a new aggregate of data comparable to the preceding one”.
surveying5 and digital technologies for representation and communication. To optimize

data captured at Villa Borghese in surveying campaigns, it was decided to combine
different surveying methods and instruments with the objective to control error in order
to be better confront problems involved in surveying architectural and sculpture ele-
ments of medium and small scale. The study of villa Borghese architecture is based on
a process that integrates traditional surveying techniques with those of low cost,
non-contact ones. This methodology makes it possible to ascertain that the direct
survey control the measurements and the image based modeling6 techniques the model
construction [3]. The experimented procedure was aimed at constructing qualitatively
controllable 3D and 2D graphic models where it is essential to define the uncertainty
level of the model scale. The surveying procedure yielded a numeric model (point
cloud) – an objective datum which avoids conscious schematization done by the
operator and is configured as the departure point for further elaborations. The quality of
the data obtained with various software which work with this technology has been
verified by attributing to each model a scale reference that describes its capacity to
reproduce reality digitally in relation to its geometric features and those related to
surface qualities. The sequence of works started with data acquisition (surveying) to
collect some sets of photographs7. It was decided at the outset to proceed with data
acquisition applying the same criteria for all the objects analyzed doing general and
detailed photo shots with 30 % of overlapping to obtain as full coverage as possible. In
order to describe the characteristics of each type of element and to define survey
accuracy, tolerances, methods and output the following step was performed for
choosing the most appropriate technique taking the following into account: object
characteristics (its dimension/volume, size threshold at micro- and macro-scale),
material characteristics; capture instrument characteristics (accuracy and resolution
required, recovery area dimensions); lighting conditions; presence or absence of lam-
bertian materials; aim of the survey (size and minimum level of detail to be returned).
Data obtained were elaborated with Agisoft Photoscan, commercial software to obtain
3D model from photographic images. Data have been processed to reconstruct a low
density point cloud that became very dense in the successive phase. Finally, 3D mesh
model of the object was obtained with an algorithm called SIFT (Scale Invariant
Feature Transform). Photo elaboration with Photoscan ensured the identification of the
object’s salient points, which made it also possible to control the uncertainty level of
metric values, always ˂3 mm. In the majority of cases we were able to obtain a
complete reconstruction of the object overcoming major problems of coverage (Fig. 3).
It becomes clear now how the data acquired with different surveying methodologies
5
Technologies that make it possible to capture millions of points on surfaces without the requiste of
establish the measured distances. The result of their application differs qualitatively and
quantitatively from that obtained with traditional instruments, which always assume the operator's
selection of the salient points to measure of the object of study (3D laser scanning, Structure from
Motion technology, Gigaphoto).
6
This technique offers the possibility to generate 3D content from photographic images without the
need of high-cost hardware. This technology originates in the theory of photogrammetry and makes
it possible to build 3D models on the basis of photographs.
7
NIKON D300 camera with CMOS sensor of 12 MP.
became comparable and indispensable for a heterogeneous information database

structure that yields an ever more detailed rendition of the object analyzed [4]. Data
elaboration belongs with the stage of survey. 3D textured models worked out were
taken as the point of departure for devising standard 2D models8. A low cost survey
conducted with quick methodologies enabled us to acquire highly flexible data which
could define the characteristic features of the object (its geometry, topology, texture) in
geometrical-proportional and architectural drawings, degradation and color analyses.
The definition of such 2D and 3D models is the result of considerations prompted by
the will to defining a standard for communicating the characteristics of all the objects
surveyed in the same manner [8]. The choice is related to the experimental aspect of the
surveying methodology adopted, considered especially suitable for artifacts of this type
(Figs. 1 and 2).
3 Database Management System
Modeling proved considerably advantageous in the stage of documentation: the output

being an informative model which lead to critical interpretations [5]. An information
system that divulges connections and links between high definition 3D models and
heterogeneous data turns out to be an extensively articulated structure [6]. It must be
able to acquire and modify spatial data in order to archive, maintain, analyze and
present them in the alphanumeric and/or graphic form. Building up an archive whose
objective is to preserve and popularize objects of CH involved some problems in
working out the conceptual model, data acquisition modalities, visualization, naviga-
tion and access to the contents [7]. The concept of information structured within a
database lies at the very centre of the present research. The Archive of Villa Borghese
architecture aims at identifying the kind of information of the architectural object and
classifying them into distinctive and characteristic categories, which make it possible to
cognize the given object. Such knowledge can be achieved only after cognizing the
precise geometry of the architectural structure in close relation to other univocal and
selective information, concern the structure, construction techniques, the state of
preservation and the materials of which the object was built. Moreover, it transcends
the metadata9 generated during the surveying campaign as well as all the documen-
tation collected so far derived from historical-architectural inquiries and surveys con-
ducted on the objects in question. One modality to popularize objects of CH has been
8
Geometric (1:100) and architectural (1:50) plans, elevations and sections.
9
Metadata, considered as data about data, can help to organize information and provide digital
identification and to understand the process through which one or more artifacts were processed or
interpreted (The London Charter, version 2.1).
Fig. 1. Cataloging of Villa Borghese architectures. From huge (Buidings) to medium

(Monumental entrances, Street furniture, Fontains) and small scale (Monuments).
Fig. 2. Cataloging and analysis of obtained results. Software Agisoft Photoscan. Case study:
Fontains.
Fig. 3. From 3D mesh model to 2D models. Case study: Edicola della Musa.
provided by software 3DHOP10 that ensures visualization, with the ultimate goal of
sharing and spreading information. It allows data entry by different cataloguers,
qualification of 3D models in relation to different levels of detail, semantic structure
and high resolution real-time visualization of models [10]. Database devising implies
an attempt at answering some questions involved in designing compartments of the
informatics system in order to guarantee its use with the passage of time; the ability to
sustain diverse formats in order to ensure data integration; connection with
10
3DHOP (3D Heritage Online Presenter) is an open-source software package for the creation of
interactive Web presentations of high-resolution 3D models, oriented to the Cultural Heritage field.
It allows the creation of interactive visualization of 3D models directly inside a standard web page,
just by adding some HTML and JavaScript components in the HTML code. The 3D scene and the
user interaction can be easily configured using a simple “declarative programming” approach, and
by a series of provided JavaScript functions (Retrieved from: http://3dhop.net/).
heterogeneous information. It integrates successfully the descriptive aspects with the

spatial nature of the real object and their formal characteristics expressed by models.
Clicking on the selected part of a model, is possible to obtain a wide gamut of
information. Special tables have been drawn up while the database fields and the
graphic interface of the web page contains a form for inserting data. These sections are
interconnected, so that it is possible to move from one section to another using simple
hyperlinks that connect the different areas (Fig. 4).
Fig. 4. Structuring digital archive to manage heterogeneous data. Case study: Edicola della
Musa.
4 Conclusions
Digital archives are considerably advantageous for disseminating information related to

CH: they can be used alongside of or complementary to physical archives; they offer
photorealistic visualizations of models of existing objects; they can be used by different
clients (specialists, generic) since it is possible to filter the contents; data are ready to be
used directly for analyses based on geometric and formal information provided by
different representation scale; it is possible to get to know the object completely
moving with ease from the general to the particular, from the simple to the complex
(and vice versa) by breaking down the objects into sub-components organized hier-
archically [9]. Taking advantage of the newest technologies, languages and commu-
nication protocols that facilitate information exchange and sharing, it is possible to
manage information on a much larger scale as well as construct instruments thanks to
which information can be really exchanged and fully shared. Today digital objects have
become a tradition in the cultural production and a business cycle with the objective to
disseminate and communicate cultural data. The management of CH is linked with the
development of new methods of acquisition and transmission of knowledge, and it
passes through an ever- increasing quantity and complexity of metadata. The appli-
cation of ICT to cultural heritage is certainly one of the most interesting frontiers of
activity since object characteristics can be extracted from 3D images and models and
can be made interactive.
References
1. Bianchini, C., Borgogni, F., Ippolito, A., Senatore, L.J., Capiato, E., Capocefalo, C.,
Cosentino, F.: From surveying to representation: theoretical background, practical issues,
possible guidelines. In: 18th International Conference on Virtual Systems and Multimedia,
VSMM 2012, pp. 507–513 (2013)
2. Bianchini, C.: L’approccio scientifico al rilevamento del teatro di Mérida. In: La
documentazione dei teatri antichi del mediterraneo, pp. 22–37. Gangemi, Roma (2012)
3. De Luca, L.: La fotomodellazione”, Flaccovio Editore, Palermo (2011)
4. Remondino, F., Campana, S.: 3D Recording and Modelling in Archaeology and Cultural
Heritage - Theory and Best Practices. Archaeopress BAR Publication Series, Oxford (2014)
5. Siotto, E., Callieri, M., Pingi, P., Scopigno, R., Benassi, L., Parri, A., La Monica, D.,
Ferrara, A.: From the archival documentation to standardised web database and 3D models:
the case study of the camaldolese abbey in Volterra (Italy). In: 18th International Conference
on Cultural Heritage and New Technologies, CHNT 18, pp. 1–9 (2014)
6. Minto, S., Remondino, F.: Online access and sharing of reality-based 3D models.
SCIRES-IT SCientific RESearch and Information Technology Ricerca Scientifica
e Tecnologie dell’Informazione 2(2), 18–27 (2014)
7. Bartolomei, C., Ippolito, A.: Low-cost cataloguing methodologies for architecture: the
system of gates in Bologna. In: Ioannides, M., Magnenat-Thalmann, N., Fink, E., Yen, A.,
Quak, E., Way, E. (eds.) Digital Heritage: Progress in Cultural Heritage, Documentation,
Preservation and Protection, EuroMed 2014, pp. 249–257. Multi-Science Publishing Co.
Ltd., United Kingdom (2014)
8. Bianchini, C., Borgogni, F., Ippolito, A., Senatore, Luca, J.: The surveying and
representation process applied to archaeology: a quest for invariants in a highly variable
context. In: Di Giamberardino, P., Iacoviello, D., Jorge, R.N., Tavares, J.M.R.S. (eds.)
Computational Modeling of Objects Presented in Images. LNCVB, vol. 15, pp. 1–29.
9. Gaiani, M., Benedetti, B., Apollonio, F.I.: Teorie per rappresentare e comunicare i siti
archeologici attraverso modelli critici. SCIRES-IT SCientific RESearch and Information
Technology Ricerca Scientifica e Tecnologie dell’Informazione 1(2), 37–70 (2011)
10. De Luca, L., Busarayat, C., Stefani, C., Renaudin, N., Florenzano, M., Véron, P.: An
iconography-based modeling approach for the spatio-temporal analysis of architectural
heritage. In: Shape Modeling International Conference (SMI 2010), Washington, USA,
pp. 78–89 (2010)
Digital 3D Reconstructed Models – Structuring
Visualisation Project Workflows
Mieke Pfarr-Harfst1 ✉ and Stefanie Wefers2

( )
1
TU Darmstadt, Darmstadt, Germany
[email protected]
2
i3mainz – Institute for Spatial Information and Surveying Technology,
Mainz University of Applied Sciences, Mainz, Germany
[email protected]
Abstract. Cultural Heritage (CH) visualisations have to be understood as a

combination of research sources, the contemporary historical and cultural context
(Zeitgeist), project background and work process. All available information is
collected, consolidated, filtered and assembled into a coherent picture. In case of
digital 3D reconstructed models, the result is a digital data set that can be
processed for different application fields. They are understood as a result of a
complex creative process and as a synthesis of a CH research project, its CH
context, the available research source material, and the modeling process itself.
For all visualisation types in CH different conditions, factors, and basic rules apply
to achieve a high quality result. Two examples are presented illustrating the
structured view on visualisation projects as such. This paper seeks to differentiate
the various research sources being the basis for digital 3D reconstructed models
and defines work phases allowing a quality assessment. Furthermore, the poten‐
tials of including this structured view into the ontology COSCHKR currently under
development is discussed. In combindation with traditional guidelines COSCHKR
platform could open up new and flexible approaches.
Keywords: Visualisation · 3D reconstruction · Cultural heritage · Workflow
1 Introduction
Visualisation is “the act or process of interpreting in visual terms or of putting into visible
form” [1]. This universal definition encompasses all kind of figurative display (digital
and physical) such as 2D images, drawings, paintings, but also 3D representations, 3D
point clouds, 3D models, and 3D print-outs. Such kind of visualisations are used within
CH disciplines to convey a specific message with respect to a research result which is
not clearly visible in the material CH itself since it is e.g. only partly preserved or lost
its context of usage. Therefore, a CH visualisation combines facts and interpretations
[2]. They are a result of an interdisciplinary, collaborative process in which mostly
architecture, archaeology, and the history of art and architecture are involved. However,
due to recently evolving digital technologies, more disciplines are involved in this
process such as computer science and surveyors.

DOI: 10.1007/978-3-319-48496-9_43
Digital 3D Reconstructed Models 545
Digital 3D reconstructed models (further on “digital models”) are used as a new

medium in the field of transfer of knowledge since about 25 years [3]. The benefit of
these data sets is that various output formats can be chosen depending on the aimed
application. In the last decade the application fields and possibilities of these digital
models are, however, under discussion: In addition to a transfer of knowledge which
very often is only linked to one event such as a film sequence shown in one exhibition,
the value of these digital reconstructed models is acknowledged through their applica‐
tion already during research and in a long-term perspective also for preservation and
documentation of condition status and processes. As such CH visualisations have to be
understood as a combination of research sources [4], the contemporary historical and
cultural context (Zeitgeist), the project background and the work process. All available
information can be collected, consolidated, filtered and assembled into a coherent
picture. In the case of digital models, the result is a digital data set that can be processed
for different application fields.
As soon as digital models are not only understood as a mere visualisation but also
as a tool documenting the generation process, the project planning and handling of all
information gets very complex. The overall aim of all projects creating a digital 3D
representation should be to enable a continuous re-use also for later projects. One essen‐
tial part for this aim is a structured view on such projects creating a conceptual basis.
Therefore, in a first attempt this paper seeks to differentiate the various research sources
being the basis for digital models and define project work allowing a quality assessment.
2 Definition of Terms
The term “visualisation” summarises in the field of CH various types of display and
application. All these types display complex relationships and aspects using a variety
of visual media ranging from simple 2D applications to digital models including draw‐
ings, 3D point clouds or meshs. Technologies such as laser scanning or SfM acquire or
generate 3D data, whereas digital models usually combine heterogeneous data sources
including 3D data representing material CH and results from hand-made, artificial
modeling processes.
We focus on digital models as they are a result of a complex creative process
including various disciplines. As soon as they are used in the context of a CH research
project or as a tool to visualise the results of such projects, they have to be understood
as a synthesis of the project, its context, the available research source material, and the
modelling process itself. For all visualisation types in CH different conditions, factors,
and basic rules apply to achieve a high quality result.
Terms such as “computer-based visualisation” and “(3D) reconstructed model” are
suggested by Daniele et al. [5] and Denard [4]. The first is defined as “The process of
representing information visually with the aid of computer technologies” [4] being a
general term for all kind of digital visualisations. And the second is defined as “A 3D
model that has been created based on the shape and appearance of a real object. The
reconstruction is done either by technological means as 3D scanning (see [Sampled
Model]) or by scientific data such as architectural designs and reference images” [5]
546 M. Pfarr-Harfst and S. Wefers
being more specific and reflecting our understanding of 3D models. It takes the need
into account to differentiate the various digital data sets which are used as a supporting
tool in the generation of a 3D model or which are integrated in the 3D model. The
“Sampled Model” is “A 3D model that has been created through means of digital scan‐
ning from a real world existing object. The scanning is done with a device that analyses
the real-world object and collects data on its shape and nowadays its appearance” [5].
However, it is also necessary to be able to name those parts of a digital 3D reconstructed
model which are not sampled models but are the result of a hand-made modelling
process. For these digital data sets we suggest the term “hand-made digital reconstruc‐
tion”.
Furthermore, we suggest adding the following more detailed definition to the above
cited definition of the term “digital (3D) reconstructed model”: digital models are
computer-based models of landscapes, buildings, building structures, material objects
and/or structural elements in which the object-based knowledge is gathered, consoli‐
dated, compacted and visualised. The consequence of this process is the visualisation
of knowledge, why these models are mirrors of recent research and an object of future
research; as such they are an innovative and future-orientated tool in the field of research,
preservation and transfer of knowledge.
3 State of the Art: Visualisations Through Digital 3D

Reconstructed Models
For more than 20 years, visualisations and especially digital models have been applied
in the field of CH research [6, 7] merging and visualising object-related knowledge.
Being a synthesis of research results, the historio-cultural context, and modelling
processes, the 3D data set is used as a basis for a variety of display modes, output and
presentation formats in various fields of application [3, 8]. 3D models are mainly used
as a visualisation of CH assets in the context of exhibitions [8]. However, they are also
present in other media representations such as TV documentaries [9] or web-based
applications [10, 11] and affect the image of the past. Through recent technological
developments it has become possible that everybody can create such models. However,
the evaluation of the scientific quality of these models often remains sidelined [12, 13].
In the current and future field of research, visualisations in the CH domain become more
and more important due to their potentials and possibilities of application [8, 14]. This
is associated with a number of challenges: issues such as the development of a meth‐
odology or process-related guideline or generally applicable strategy and the documen‐
tation of the underlying resources and the processes in terms of a scientific justification
are present in the current discussion. For digital reconstructed models no solutions are
defined, yet [15]. Current topics under discussion on a national and international level
are basic principles, methodology, documentation, knowledge representation, and
appearance; in Germany a working group “Digital reconstructions” was founded in
2013. The idea is to create a community network on a national level giving the basis for
a common understanding on the above mentioned topics.
Two of the above mentioned aspects of visualisation in CH will be discussed in this

paper: methodology and documentation. In the area of methodology theoretical studies
have been previously carried out. One of the most important results of these studies is,
that goal, intention, and application context always belong together and affect the process
and result of a visualisation project [3, 8, 16]. First ideas of guidelines are provided by
the London and Seville Charter [17].
In the last five years, knowledge evidence of 3D data sets with respect to the devel‐
opment of documentation strategies has become more and more relevant. Projects such
as “coloniaI3D” [18], “Bamberg 4D” [19], and “MayaArch3D” [20] as well as the still
ongoing project “Virtual reconstructions in transnational research environments - The
portal: palaces and parks in the former East Prussia” [21] and “Toporaz topography in
space and time” [22] deliver first approaches to the establishment of spatial virtual plat‐
forms combined with the annotation of research sources and process data to the 3D
models. For quality assurance it is also necessary to develop a system assisting the
reconstruction process. The aspect of usability is one of the most important ones in this
context. The system to be developed must be practicable, easy to understand, and devel‐
oped on the basis of the user needs. Other projects are dedicated to the theoretical
discussion of this topic as a basis for the development of a practical documentation
system [23, 24].
In 2010, a 4-level system was developed as a documentation strategy, which reflects
the complexity of the process and the basics of scientific visualisations [12]: Level 1 is
the background of the project which is, the knowledge of the models: project partners,
intention, technology, results; level 2 contains the project context, the knowledge that
is represented by the models, aside from the knowledge about the models; level 3 defines
the data structure including all available research sources and their nomenclature. This
should be done individually as every project has its own specifications; level 4 describes
the connection of all research sources to the reconstructed model and the final product
of a modelling process. The essential point is to document both research sources and
hand-made reconstructed models as well as the sampled model. Main decisions and
milestones should be depicted. Starting point is the text-based construction description,
an overview of the key data on the building with cross references to the so-called source
and methods catalogues. In a building’s research source catalogue, the research source
is assigned directly to a specific object or building of the entire visualisation project, and
in the methods catalogue the project is linked to the research sources and the process.
This 4-level system reflects a documentation strategy saving the knowledge contained
in a CH visualisation project and additionally its modelling process by archiving e.g.
interim modelling results.
An important contribution to the theoretical discussion can be found in Bentkowska-
Kafel et al. [25]: two necessary aspects “transparency” and “paradata” are discussed in
the context of quality assurance of CH visualisation. Both terms focus on evidence of
knowledge, sustainability of data, but also the presentation of hypotheses and findings.
Nevertheless, the scientific debate regarding scientific quality assurance of visualisa‐
tions in CH has just started.
4 Structuring the Creation of Digital 3D Reconstructed Models
CH visualisations, especially digital models, and their creation process are influenced
by a high number of factors. Analysing 19 CH 3D visualisation projects of three different
institutions (King´s College London, University of Sarajevo, Technische Universität
Darmstadt), it is possible to define a first framework of such a working process through
the identification of similarities and differences of typical processes and methodolo‐
gies [26].
The following research questions were the basis for the investigation: Are there any
similarities between the different workflows and methods? What are the similarities
(definition), what are the differences (definition)? What are the reasons for the differ‐
ences and the similarities relating to context of the 3D model? Are there any depend‐
encies between properties, potentials, and application scope and possibilities? Is it
possible to define some general phases of a workflow?
Together with results from earlier studies [12] the following factors influencing the
process of 3D modelling were defined: background of the project; context of the project;
timeline; involved people; intention/aims of the project (research, transfer of knowledge,
preservation); application field; application/preservation format; application possibility;
type of 3D visualisation method; technical system/aspects; methodology and steps of
processes; results.
Based on this, it was possible to define a framework of a workflow. It consists of
four main phases and the background of the project as a general frame. The four project
phases are “preparation”, “data acquisition”, “data processing”, and “finishing” [26].
All parameters such as involved people, timeline, intention, expected result, financial
condition as well as available research sources and technical systems, which are neces‐
sary realising a project, are fixed within the preparation phase. The output of this phase
generates the input for the phase “data acquisition” affecting the methods and techniques
for generating the data.
The input of the phase “data processing” is based on the results of “data acquisition”
and the selected method. Within the 3rd phase the acquired data are used in various ways
depending on the project intention and purpose. Both phases are therefore linked as the
quality and content of the data sets acquired in the 2nd phase have to meet the require‐
ments of the phase “data processing”. Through harmonisation of both phases the entire
project benefits. The result of this phase is a digital data set (2D or 3D), which is the
basis for the “finishing” phase providing data formats which serve the needs of the
application defined by the project aim. Various application fields and possibilities
require a great scope of processing the digital data set.
Within and in-between the phases an input-output principle exist determining the
next step and the way how work packages (WP) as part of the phases interact (iterative
or linear). Actually, this depends on the project background such as the purpose,
involved people, available techniques, timeline, and funding. Therefore, the background
of the project generates the frame and working process.
Three main milestones exist: The first one is the defined background of the project
being the basis and input for all phases and WPs. The second is reached when all neces‐
sary data are acquired being an input for the phase “data processing”. The third milestone
is the digital data set - the digital reconstructed model - available after the data processing
phase being the input for the finishing phase creating the result. At the beginning of a
project (within preparation phase) it is necessary to define further milestones depending
on individual parameters.
The complexity of the work process is particularly evident in the phase “data
processing”. As examples, the projects “Body and Mask in ancient Space” [27] and
“Oplontis Visualisation Project” [28] are described in more detail. The focus of the first
project was identifying conditions and actualities of ancient theatre as well as the roll
of body and masks in the context of theatres using digital visualisation methods. Within
the interdisciplinary collaborative research project, different digital techniques such as
3D scanning, 3D modelling, 3D motion capturing, and rapid prototyping were used. The
following different CH visualisation types were applied in a virtual research environ‐
ment: film sequences, hand-made 3D reconstructions (theatre), sampled models of
objects (masks) and characters (actors), 3D point clouds acquired with laser scanners.
The process consisted of eight different WPs connected in various ways not following
a linear order. It was rather the case that the input or output of a WP fixed single steps
bringing out a non-linear working process (Fig. 1).
Fig. 1. Working process of project “Body and Mask”.
It becomes clear that the various output data that were generated in the phase of “data
acquisition” were incorporated in various WPs within the phase “data processing”. The
digital reconstruction of the theatre as a setting of the subsequent performances of the
actors has been resorted to an existing digital 3D model of an ancient theatre. Further‐
more, digital 3D models of masks were generated using a 3D laser scanner. The digital
masks were again transferred into a physical format through 3D print-outs. The 3D print-
outs of the masks were coloured and used by the actors in the simulation of the perform‐
ance. The performance was recorded by the motion capture process and inserted into
the digital 3D model of the theatre.
The entire digital 3D reconstructed model was used to generate various output
formats such as film sequences and 2D renderings (finishing phase). This process is
characterised within “Data Processing” of the further processing of data in several ways.
The WPs were linked in an iterative and linear process, too.
The main issue of the Oplontis project was to bring all research results and biblio‐
graphic references together in an interactive, real-time 3D model. This model should be
used as a virtual research environment, a documentation of the actual status quo of the
research and a systematic multidisciplinary study of Villa A and Villa B at Oplontis.
Research questions were e.g. the chronological understanding of the villas, the relation
of building and ancient landscape, the function of the villas as well as the social structure
of the inhabitants. The following different CH visualisation types were used in the virtual
research environment: hand-made 3D reconstructed models (villa), sampled models (3D
point clouds).
The workflow of the project Oplontis was non-linear, iterative, and intersecting.
Interlinking the single WPs is a very popular workflow (Fig. 2). However, it is also
noticeable that within the phase “data processing” the results of different WPs are always
a basis for other WPs.
Preparation
Data Collecting
Data Processing
Publishing/Finishing
3D modelling
Background of the project
Digital Dataset
Collecting
Definition of
Material 3D Archive
discussion
3D
(digital + verifying
non-digital) 3D VRE
texturing
lightning
Timeline of the project
Fig. 2. Working process of the “Oplontis” project.
For most dissemination projects creating digital models, a circular interaction

between WPs is observed. Additionally, within the phase “data processing” four tasks
can be defined: basic research, 3D modelling, verification, and texturing. The basic
research is not limited to the phase of project preparation; in fact it accompanies the
entire process. This includes not only the classic source research and evaluation, but
also the preparation and processing of digital output data such as 3D point clouds gener‐
ated e.g. by laser scanners. The 3D modelling together with the verification forms the
core of the actual process. A decisive factor is that the 3D modelling needs to be under‐
stood as a craft, which is dependent on the modeller, the technical system, and the basics.
Through verification the quality in scholarly and content sense of the digital 3D recon‐
struced model is assured. Additionally, a need for quality assurance within the data
processing with respect to technical aspects exists. The texturing usually happens at the
end of the modelling process of each singular model being part of an entire 3D recon‐
structed model. Therefore, the texturing task is interlinked with all WPs.
Between the tasks an input-output principle exists in all directions. The available
sources (ranging from 3D point clouds to drawings) are e.g. input for the 3D modelling
which produces as output a 3D model. This output in turn is used as input for the
verification or texturing task. The output of the verification is a result which in turn can
serve as input for the 3D modelling or for basic research.
5 Discussion
By structuring visualisation projects as described above, factors having impact on the

entire workflow, content or output are made visible. A more detailed structuring will
support the project planning as knowledge gaps between various involved experts will
be bridged. Especially the phase “data acquisition” encompassing the digitisation of
maps, ground plots etc. as well as the spectral and/or spatial recording of physical objects
allows a workflow optimisation as a deterministic connection between the physical
characteristics of the object, the capabilities of the recording device and the required
data content and quality exists.
Currently, the so-called COSCHKR platform is under development which is intended
to give recommendations for spatial and spectral recording strategies with respect to CH
objects, CH applications, external influences and required data [29–31]. The COST
Action TD1201: “Colour and Space in Cultural Heritage (COSCH)” gives the oppor‐
tunity to develop such kind of a platform structuring and linking the knowledge of the
various disciplines involved in the recording of CH objects as through this Action a
European network exists bringing all experts together which are necessary for such kind
of a discussion and structuring (conservators, CH experts, engineers, surveyors, IT
experts, etc.). The platform envelops an ontology so-called COSCHKR (Knowledge
Representation; for technical details see [29, 30]) which is expressed with Web Ontology
Language (OWL) [32]. It is a domain ontology with hierarchically structured classes
which are interlinked through rules defining dependencies of the technical parameters
(such as object size, measurement accuracy etc.) required to set-up a reliable 3D or
spectral digitisation strategy for a specific CH application. Due to this intention this
domain ontology is completely different to formal ontology CIDOC-CRM developed
for archives and museums. It enables the structuring of all biographical information
about CH (e.g. provenance). The entire ontology is built-up upon the following five top-
level classes: Technologies, Data, CH Application, Physical Thing, and External Influ‐
ences (Fig. 3). The sub-classes of Technology were created through a global view on
expert knowledge [33]. Through logical rules and dependencies a link is established
between CH Application and Data requirements which allows exploiting the ontology
and retrieving recommendations.
For the definition of the class “Physical Thing” we refer to the CIDOC-CRM class
E18 “Physical Thing”: “This class comprises all persistent physical items with a rela‐
tively stable form, man-made or natural. […]” [34]. The subclasses and characteristics
which define the “Physical Thing” within COSCHKR ontology in more detail are all
Fig. 3. Top-level classes of the ontology enveloped by the COSCHKR platform. The classes are
linked through rules (arrows) and dependencies.
technology oriented. They describe the physical appearance and shape of a physical
thing and do not refer to terms describing their function or purpose in the real world
(such as church, fibula, ceramic etc.). To give a short impression on the structure, e.g.
the death mask of Tutankhamun would be characterised as a small to medium sized
“Physical Thing” (class “Size”) with a complex 3D shape (classes “Structure” and
“Shape”) and a partly highly reflective (gold; class “Reflectance”) and partly translucent
(glass and precious stone inlays; class “Transparency”) surface. With respect to a visu‐
alisation project the “Physical Thing” could be a ground plot or the ruins of a building.
However, the object which should be digitised has to be put in the context of the “CH
Application” and its requirements on the “Data”. Each “CH Applications” requires
different data quality and content (Fig. 3), e.g. a CH application requires a high spatial
accuracy or a high spectral quality. Therefore, the “CH Application” provides a state‐
ment of requirements (a) with respect to the “Physical Thing” and (b) with respect to
“Data” which are needed for the “CH Application”.
“A manner of accomplishing a task especially using technical processes, methods,
or knowledge” [35] defines the scope of the class “Technologies” (Fig. 3). Through its
subclasses, dependencies, and rules the acquisition and processing of the “Data” with
tools required by the “CH Application” are described. Since the ontology so far focuses
on spectral and spatial recording and data usage, the classes contain information about
spectral and spatial devices, data acquisition and processing. For visualisation projects
further digitisation devices including the description of their capabilities would need to
be integrated. The generated “Data” (Fig. 3) are digital representations of the “Physical
Thing”. Its physical characteristics determine the choice of the recording device as they
have major influence on the data quality. Furthermore, the acquisition and processing
of data is affected through “External Influences” (Fig. 3) such as site illumination, visi‐
bility or staff competence.
The scope of the class “Data” (Fig. 3) is understood as “facts or information used
usually to calculate, analyse, or plan something” [36]. Its subclasses describe all kind
of acquired or processed data including data about the “Physical Thing”, e.g. published
in a book. As the class “Technologies” also might have demands on the class “Data” it
is a 2-way relationship (Fig. 3): E.g. for the generation of digital reconstructed models
all kind of source materials might play a role and have to be included in the data
processing.
Through the 5 top-level classes all aspects relevant within the two phases of a digi‐
tisation and visualisation (data acquisition and processing triggered by the finishing
phase) are covered. The intention is to implement mainly the phases “data acquisition”
and “data processing” of visualisation projects in the above described ontology
(COSCHKR). The essential part is to define and structure the requirements of the phase
“data processing” on the acquired data within visualisation projects. E.g. a 3D point
cloud used as basis for the spatial dimension of a hand-made reconstructed model does
not need a high point density, whereas a high point density with colour information
would be needed if the intention would be to integrate the 3D point cloud in the digital
3D reconstructed model. Finally, the COSCHKR platform will be able to give recom‐
mendations for the set-up of visualisation projects. On the one hand it will be possible
to check if available data sets meet the requirements of the data processing and appli‐
cation. On the other hand it will support the data acquisition (especially 3D data and
spectral data), which are satisfying the project’s needs.
6 Conclusion
Through the investigation of 19 visualisation projects the interaction between WPs was
identified and structured as linear, iterative or both. The differences are a result of the
project background. The circular, alternating process is best suitable for projects with a
complex content or intention and uncertain research sources. Often questions or queries
occur during complex 3D modelling projects that can be answered only in the course of
the project in cooperation with all partners. Therefore, it is recommended to use a linear
process for less complex digital models containing an easy manageable content with a
clear project intention. At the beginning of a project it should be decided on the basis
of all available facts and parameters which working process fits best. In the future, the
project planning phase could be supported not only through traditional guidelines but
also through the above suggested COSCHKR platform.
Acknowledgement. This work was partly supported by COST under Action TD1201: Colour
and Space in Cultural Heritage (COSCH). Furthermore, we would like to thank Prof. Dr.-Ing.
Frank Boochs and Dr. Ashish Karmacharya (i3mainz) for discussions and support.
References
1. Merriam-Webster. http://www.merriam-webster.com/dictionary/visualization
2. Rizvić, S., Tsiafaki, D.: Interim Report 2012–2014. COSCH WG5: Visualisation of CH
Objects and its Dissemination. COSCH e-Bulletin 3, Mainz (2016)
3. Pfarr-Harfst, M.: 25 years of experience in virtual reconstructions - research projects, status
quo of current research and visions for the future. In: Giligny, F., Djindjian, F., Costa, L.,
Robert, S. (eds.) CAA 2014: 21st Century Archaeology: Concepts, Methods and Tools.
Proceedings of the 42nd Annual Conference on Computer Applications and Quantitative
Methods in Archaeology, pp. 585–592. Archaeopress, Oxford (2015)
4. Denard, H.: The London Charter for the computer-based Visualisation of cultural heritage.
Draft 2.1, 7 February 2009. http://www.londoncharter.org/
5. Daniele, F., Alfonsina, P., Mohamed, F.: Terminology, definitions and types of virtual
museums. V-must Experience the future of the past. Deliverable report D2.1c (2014). http://
www.v-must.net/library
6. Grellert, M., Koob, M., Pfarr, M.: Ephesos - Byzantinisches Erbe des Abendlandes. Digitale
Simulation und Rekonstruktion der Stadt Ephesos im 6. Jahrhundert. In: Daim, F., Drauschke,
J. (eds.) Byzanz - das Römerreich im Mittelalter: Pracht im Alltag eines Weltreiches. Teil 2,2,
Schauplätze, pp. 731–744. Verlag des RGZM, Mainz (2010)
7. Pfarr-Harfst, M., Grellert, M.: 25 years virtual reconstructions - actual challenges and the
comeback of physical models. In: Proceedings of Digital Heritage International Congress, pp.
91–94. IEEE Press, New York (2013)
8. Münster, S.: Entstehungs- und Verwendungskontexte von 3D-CAD-Modellen in den
Geschichtswissenschaften. In: Meissner, K., Engelien, M. (eds.) Virtual Enterprises,
Communities and Social Networks, pp. 99–108. TUDpress, Dresden (2011)
9. Unsere Geschichte: Ostpreußens vergessene Schlösser, project description. www.ndr.de/-
fernsehen/sendungen/unsere_geschichte/Unsere-Geschichte-Ostpreussens-vergessene-
Schloesser,sendung514286.html and video: www.youtube.com/watch?v=Pxe7YgAEs2w
10. Forschungsarbeiten des RGZM in Ephesos. www.rgzm.de/ephesos/
11. DHOP – 3D Heritage Online Presenter. http://3dhop.net/
12. Pfarr, M.: Dokumentationssystem für digitale Rekonstruktionen am Beispiel der Grabanlage
Zhaoling, Provinz Shaanxi, China. TUprints, Darmstadt (2010)
13. Dwarehouse. https://3dwarehouse.sketchup.com/?hl=de
14. Grellert, M.: Immaterielle Zeugnisse – Synagogen in Deutschland: Potentiale digitaler
Technologien für das Erinnern zerstörter Architektur. Transcript-Verlag, Bielefeld (2007)
15. Münster, S., Kuroczynski, P., Pfarr-Harfst, M., Lengyel, D.: Future research challenges for a
computer-based interpretative 3D reconstruction of cultural heritage - a german community’s
view. In: Proceedings of 25th International CIPA Symposium 2015 on SPRS Annals of the
Photogrammetry, Remote Sensing and Spatial Information Sciences, 31 August – 04
September 2015, Taipei, vol. II-5/W3, pp. 207–2013 (2015)
16. Münster, S.: Interdisziplinäre Kooperation bei der Erstellung virtueller geschichtswiss
enschaftlicher 3D-Rekonstruktionen. Springer, Heidelberg (2016)
17. Lopez-Menchero, V., Grande, A.: The principles of the Seville Charter. In: Proceedings of
the XXIII CIPA Symposium, Prague, Czech Republic, 12–16 September 2011. http://
cipa.icomos.org/fileadmin/template/doc/PRAGUE/096.pdf
18. Colonia 3D. http://colonia3d.de/colonia3d-home/
19. Bamberg4D. https://www.uni-bamberg.de/bauforschung/forschung/projekte/digitales-
stadtmodell/4d/
20. MayaArch3D. http://www.mayaarch3d.org/language/en/sample-page/
21. Virtual reconstructions in transnational research environments - The portal: palaces and parks
in the former East Prussia. http://www.herder-institut.de/index.php?id=585
22. Toporaz topography in space and time. https://www.fiz-karlsruhe.de/de/forschung/projekte/
toporaz-topographie-in-raum-und-zeit.html
23. Kuroczyński, P., Hauck, O., Dworak, D.: Digital reconstruction of cultural heritage –
questions of documentation and visualisation standards for 3D content. In: Ioannides, M.,
Mageanat-Thalmann, N., Fink, E., Zarnic, R., Yen, A., Quak, E. (eds.) Proceedings of 5th
International Conference on Digital Heritage. Progress in Cultural Heritage: Documentation,
Preservation and Protection, EuroMed 2014, Limassol, Cyprus, 3–8 November 2014,
Springer, Heidelberg (2014). http://www.academia.edu/9189049/
24. Bruschke, J., Wacker, M.: Neuartige Werkzeuge für die Entwicklung und Dokumentation
digitaler Rekonstruktionen. In: TU Dresden Forschungsförderung und Transfer,
TechnologieZentrumDresden GmbH, Industrie- und Handelskammer Dresden, GWT-TUD
GmbH (eds.) Dresdner Transferbrief, p. 9. TUDpress, Dresden (2015)
25. Bentkowska-Kafel, A., Denard, H., Baker, D. (eds.): Paradata and Transparency in Virtual
Heritage. Ashgate, Farnham (2012)
26. Pfarr-Harfst, M.: Evaluation of selected projects at the Faculty of Electrical Engineering
Sarajevo and King’s College London. In: Rizvić, S., Tsiafaki, D. (eds.) Interim report 2012–
2014. COSCH WG5: Visualisation of CH Objects and its Dissemination. COSCH e-Bulletin
3, Mainz (2016). http://cosch.info/bulletin-3-2015
27. The Body and Mask in Ancient Theatre Space. http://www.kvl.cch.kcl.ac.uk/masks/
28. The Oplontis Visualization Project. http://www.oplontisproject.org/index.php/people/the-
oplontis-visualization-project/
29. Boochs, F., Bentkowska-Kafel, A., Degrigny, C., Hauta-Kasari, M., Rizvic, S., Trémeau, A.:
Towards optimal spectral and spatial documentation of cultural heritage. COSCH – an
interdisciplinary action in the COST framework. In: Proceedings of XXIV International CIPA
Symposium 2013 on Recording, Documentation and Cooperation for Cultural Heritage.
ISPRS Archives, Strasbourg, France, 2–6 September 2013, vol. XL-5/W2, pp. 109–113.
Curran Associates, Inc., Red Hook (2013)
30. Wefers, S., Karmacharya, A., Boochs, F.: I need high-quality data of my cultural heritage
object – Development of a platform recommending 3D and spectral digitization strategies. In:
Proceedings of the 8th International Congress on Archaeology, Computer Graphics, Cultural
Heritage and Innovation ‘ARQUEOLÓGICA 2.0’ in Valencia (Spain), 5–7 September (2016,
in press)
31. COST Action TD1201 COSCH. www.cosch.info/coschkr
32. OWL. https://www.w3.org/2001/sw/wiki/OWL
33. Wiemann, A.-K., Boochs, F., Karmacharya, A., Wefers, S.: Characterisation of spatial
techniques for optimised use in cultural heritage documentation. In: Ioannides, M., Magnenat-
Thalmann, N., Fink, E., Žarnić, R., Yen, A.-Y., Quak, E. (eds.) EuroMed 2014. LNCS, vol.
34. Le Boeuf, P., Doerr, M., Ore, C.E., Stead, S.: Definition of the CIDOC Conceptual Reference
Model. Produced by the ICOM/CIDOC Documentation Standards Group, Continued by the
CIDOC CRM Special Interest Group. Version 6.2, May 2015. http://new.cidoc-crm.org/
Version/version-6.2
35. Merriam-Webster, technology. http://www.merriam-webster.com/dictionary/technology
36. Merriam-Webster, data. http://www.merriam-webster.com/dictionary/data
Reconstruction of Wooden “Polish Manor”
Anna Rozanska1 ✉ and Wojciech Koryciński2

( )
1
Wood Technology Department,
Warsaw University of Life Sciences WULS-SGGW, Warsaw, Poland
[email protected]
2
Department of Industrial and Medicinal Plants,
University of Life Sciences in Lublin, Lublin, Poland
[email protected]
Abstract. Historical manor houses form an important part of Polish cultural

heritage. Manors referred to one storey bricked or, more frequently, wooden
buildings. The aim of this study consists in a historical, stylistic, structural and
functional analysis of Polish manors and in the reconstruction design of a wooden
manor house forming part of a homestead in the Museum of Folk Culture in
Kolbuszowa.
Keywords: Reconstruction · Manor house · Decorative wooden floors · Parquet
1 Research Aim and Background
The term “cultural substance” often appears in the context of conservation and restora‐
tion. In accordance with the definition, it refers to what is constant and remains the same,
independently of the changes in its features or external transformations. The term
“cultural substance” refers fully to architecture, including the architecture of Polish
wooden manor houses. Historical manor houses form an important part of Polish cultural
heritage. A publication dated 1659, entitled “A Short Treatise on the Construction of
Polish Manor Houses, Palaces and Castles” describes the characteristic features of Polish
architecture, taking into account three kinds of residential buildings: manor houses,
palaces and castles. Manors referred to one storey bricked or, more frequently, wooden
buildings.
The aim of this study consists in a historical, stylistic, structural and functional anal‐
ysis of Polish manors and in the reconstruction design of a wooden manor house forming
part of a homestead in the Museum of Folk Culture in Kolbuszowa.
2 Methods
The restoration design consisted in creating a model (a scheme) for the reconstruction
of the manor house. It had to take into account the historic patterns and the typical
structures, material characteristics, usage functions, conservation requirements and
technologies of manufacture. The reconstruction was based on a systematic method of

DOI: 10.1007/978-3-319-48496-9_44
Reconstruction of Wooden “Polish Manor” 557
industrial design, following the phases of designing: the analytical phase including
programming and data collection (observation, measurements, inductive reasoning),
creative phase including analysis, synthesis and preparation (evaluation, judgement,
deductive reasoning and decision) and the phase of implementation understood as
communication (including descriptions, explanations and transfer) [1]. The restoration
design was based on a historical, stylistic, structural and functional analysis of the Polish
manor house, taking into account the scope of reconstruction and research concerning
the patterns, structures and technologies of manufacture of the manor houses in South-
Eastern Poland, both preserved and not.
3 Research Results
3.1 “Polish Manor” as a Cultural and Architectural Phenomenon
A manor homestead is a closed and logical functional layout of elements that are mutu‐
ally related in view of space and composition. It includes the manor house itself with
the access road, outbuildings and other residential structures, as well as farm buildings
that often went together with ponds, a network of roads and vegetation: decorative
gardens, crops and parks. It also included fences or other elements connecting it with
its immediate surroundings. A manor homestead forms an urbanistic whole.
There were two basic types of manor homesteads that developed in Polish lands from
the second half of the 16th century till the 20th century. The first one is the practical
type, resembling an agricultural farm, with its elements grouped around the central yard.
This type, popular since the second half of the 16th century, in the baroque period,
limited the domination of manors in the composition. It was transformed under the
influence of foreign architectural designs, secular culture and lifestyle, pursuing a more
stately, sumptuous appearance and following fashionable trends. Manor homestead
transformation was also due to their demilitarisation. The new type was characterised
by the domination of the manor house and an axial layout. When feudal castles became
obsolete in the 17th century, manor houses achieved a dominant position and started to
look similar to palace layouts, just at a smaller scale (e.g. the representative, residential
part was reduced to one storey). The type of classicistic manors for middle-income and
poorer nobility was developed in Poland in the 4th quarter of the 18th century. Its char‐
acteristic features include a rectangular plan without protruding corner rooms, a porch
or portico on the axis of the front facade supported by 4 pillars, and a high roof - often
a monitor roof.
The great popularity of the Polish manor house as an architectural model is also
related to its universal functionality. The Polish manor is a residential building fit for a
wide spectrum of users. It is functional - with one storey, often built of wood on a
rectangular plan, easy to make, open to modifications or expansions.
3.2 Development of Manor Construction in South-Eastern Poland
The disappearance of defensive elements in the architecture of late 17th and early 18th
century was associated with the devaluation of old military and fortification methods;
558 A. Rozanska and W. Koryciński
and in the area of Sandomierz Wilderness, interesting for us, it was also due to the end
of Tatar raids. The development of socage estates at that time meant that many new
buildings had to be erected - manor houses, outbuildings, administrative buildings and
the so called folwarks. They were built on a rectangular plan out of wood - a resource
that was easily accessible in Sandomierz Wilderness - and their roofs were covered with
shakes. In South-Eastern regions of Poland, 80 % of buildings were built of wood. The
manor houses in that part of the country had a hall in the centre (preceded by a porch),
whose floor was usually made of bricks and walls covered with lime, while the ceilings
were made of sawn lumber or linen. On both sides of the hall there were guest rooms,
bedrooms, studies, closets, chambers and oriel rooms (protruding corner rooms). Floors
in the rooms were usually made of sawn timber, similar to the ceilings that sometimes
were covered with lime. Sometimes there were also linen ceilings in these rooms. They
were heated with tile stoves or fireplaces. The timber for doors and windows was cut
manually into logs and rather was not planed, and only the boards were sawn in the
sawmills. Windows and doors had single wooden wings mounted on hinges or pivots
and closed with locks. The outbuildings were constructed in a similar but more humble
way. The folwarks (serfdom-based farm estates) also included humble residential build‐
ings with 1 or 2 rooms and a hall, where the servants lived.
3.3 Scope of Reconstruction of Kolbuszowa Manor Homestead

In the 18th and 19th century, in the area of Sandomierz Wilderness and the contemporary
powiats: kolbuszowski, leżajski, mielecki, niżański and tarnobrzeski, manor homesteads
existed in: Bielice, Chorzelów, Dzikowiec, Kiełkowo, Kolbuszowa, Niwiska, Piskoro‐
wice, Sadkowa Góra, Trześnia near Tarnobrzeg, Trześnia near Niwiska, Trzęsówka and
Wilcza Wola. In Brzeziny, two manor houses were documented with drawings, while
the third one (not included in the documentation, so probably the least interesting as to
its architecture) has been preserved until today. Those manor houses were used as the
basis for reconstruction of a manor homestead that is currently under way in the Museum
of Folk Culture in Kolbuszowa.
The scope of manor homestead reconstruction includes the manor house with its
front, stately part and a small decorative garden, together with the part pertaining to the
farm and the administrator’s house (moved from the manor homestead in Rudna Wielka)
as well as the granary (moved from the manor homestead in Bidziny) [2]. The plan of
the estate had to admit the possibility of further expansion with such buildings as: stable
and carriage building, smithy, cowshed, brewery, etc. The design was prepared in three
versions made around the main element - a manor house based on the existing manor
from Brzeziny. Only the translocation and reconstruction of the manor house assumes
that a real building will be recreated. The entire estate is a compilation of elements and
features characteristic for manor homesteads, adapted to local conditions. The part
pertaining to the farm will be shown through the relocated: farm manager’s house,
granary and in the future other buildings, grouped around the yard. This part is connected
to the manor part through a crop garden. The manor part has an axial layout (that was
popular in the 18th and 19th centuries independently of the region) with an access
avenue, a round lawn surrounded by the driveway, the manor house, a French garden
with symmetrical planting beds and a less formal garden.
3.4 Description of the Reconstructed Manor House

19th century publications contain detailed descriptions of countryside cottages of that
region. Cottage layouts are discussed in detail together with the materials of manufac‐
ture, the types of foundations, floors, wall structures, wall protection, ceilings, kinds of
doors and windows, roof structures and covering materials, galleries, porches and stoves.
They also enumerate and describe non-residential buildings such as: barns, granaries,
stables, cellars, lumber mills, carriage houses, horse mills, sheds, pigsties, chicken
houses, dovecotes, wells and windmills. We can also find there synthetic descriptions
of the settlements on these lands, as well as plans of local farms. The topic of decorations
and ornaments is also discussed there, including flower gardens, orchards and fences
around the cottages. Nonetheless, 19th century texts do not contain any information
about the manor homesteads in that region, although sometimes manor facades, cross
sections and plans are published, suggesting large building volumes.
We can only deduct from those publications that Polish manor houses were usually
built of pine or fir wood, using the popular log house technique. Especially sumptuous
buildings were made of larch. On the corners, the logs were connected with interlocking
notch joints. After World War I, dovetail joints started to appear, perfect for smooth
walls required to apply boarding, which was very popular at that time. The posts at log
connections in very long walls had grooves, similarly to the vertical jambs of doors and
windows.
The wood used in building construction in Brzeziny was taken from the Brzezinski
forest and its diameter could reach 20 in. (12 in. on average). It was used in the form of
logs with round or square section, cut with axes or hatchets due to the better durability
of those tools in comparison with saws. Only boards and planks were sawn. Manor
houses, similarly to cottages, could be built on foundations dig in the ground and made
of very thick trees that served as support for the first log of each wall, much thicker than
the other ones, creating the frame of the sill plate. In order to protect the building against
sinking into the ground, oak rootstocks with bark (but also pine or fir saturated with
resin) or stones were placed in the ground on hard rock under the house corners. In this
case, the sill plate had to be replaced every couple of years. This is why, due to swampy
or at least wet ground, proper foundations started to be built some time later. At the
beginning, they were made of stones taken from the fields that were simply loosely
placed under the sill plates, while the corners were supported by rootstocks dug into the
ground. The usage of stones was limited due to local superstitions and the fact that
rootstocks gave better stability in boggy areas. Also stone and brick foundations started
to appear, made of irregular stones bound with mortar with one layer of bricks placed
on them (and two layers under the gable walls). The bricks, connected with mortar, were
placed in two layers, with their narrower and longer surfaces facing outside, and the sill
plate was placed directly on top of them. Under the corners, bigger stones or locally
produced cement blocks were placed. The third type of foundation was made of regularly
shaped mortar bound stones. On top of them a layer of bricks was often placed, and on
them - rarely - tar paper insulation. In houses built on slopes such high stone and brick
foundations were used as cellars. Countryside cottages usually did not have floors, only
logs longitudinally cut in halves in the part of the dwelling room devoted for cattle. The
first floors appeared in those buildings at the end of the 19th century and were made of
planed, unpainted planks placed on round beams with their top side cut, lying directly
on scattered sand. In manor houses, floors were made of beams covered with planks,
while in more stately (richer) buildings sometimes panel parquets could be found already
at the end of the 18th and beginning of the 19th century.
The manor house from Brzeziny is a wooden, one storey building with a non-habit‐
able attic, hip roof with jerkinheads, without cellar and built on a bricked foundation
(Fig. 1). It was built using the log house technique, with dovetail corners, on the plan of
a rectangle (dimensions: 21.75 m × 17.63 m), porch on the front and veranda/gallery on
the garden side. The building is quite big, with seven axes and two rows of rooms. On
each of its longer walls there are six-seven high, two-wing, symmetrically placed
windows with jamb structure. At the shorter walls on the garden side with two two-
wings windows, foundations of protruding corner rooms were found. The corner rooms
were accessed by one-wing doors that have been preserved only on the side of the
northern current kitchen. Each window wing is divided with horizontal structural battens
(muntins) into six square glass panes.
Fig. 1. Brzeziny manor house overview [3]
The section of the manor house shows the structural features of the building. Both
external and internal walls have log structure, while the core of the manor (the chimneys
and the central part of internal walls) is bricked (Fig. 2). The wall is connected with the
log structure by gradually sliding on the subsequent logs. Wooden parquets were placed
on joists lying on the sand. The beams supported boarding or very decorative wooden
panel parquets (of the Versailles type), whose patterns developed together with new
developments in interior design. Wooden ceilings are made of planks placed on top of
the beams that, in all the rooms, are placed transversally to the longer side of the
rectangle, supported by binding joists. The visible ceiling beams are decorated with
chamfers. The ground floor is 3 m high (from floor to ceiling) and the attic 4.75 m. The
hip roof with jerkinheads, with two-tier collar beam structure and crown posts, was
probably covered with shakes. This is deducted from dense roofing battens (used to
install the boarding) on the rafters. The rafters had sprockets supported by the protruding
ends of ceiling beams that changed the roof slope angle and elongated the eaves that
protected walls against water stains. The porch was covered with a saddleback roof with
boarding on the gable, and the gallery covering continued the line of the roof slope.
Fig. 2. Functional layout (a) and section (b) of Brzeziny manor house [3]
3.5 Room Functionality
The above-mentioned two-wing entrance door leads to a spacious Hall (Fig. 2). Another
two-wing doors lead from the Hall to side rooms of the front row of rooms on the left
and on the right. On the main axis, there is a two-wing door leading to the most stately
room of the manor, which could be a dining room or a living room. From the Dining
Room one may pass through a two-wing door to a side room of the garden row on the
left, and on the right to the kitchen that was recently installed there. The side rooms of
the front row and the garden row can be connected with one another.
The functional layout of the manor house is clear and logical. It is an enfilade layout
with two rows of rooms. The Hall is in the middle and behind it there is the dining/living
room - the centre of life in the manor. The rooms placed on the axis that starts with the
porch lead us further to side rooms.
Two doors in each room facilitate the circulation of traffic inside the manor house.
Initially the protruding corner rooms connected on both sides to the side rooms of the
garden row probably contained bedrooms. Their remains can be seen in the shape of the
foundations of the corner room and the door opening closed with bricks on the Southern
wall, as well as the one-wing door opening that has been preserved on the Northern side.
3.6 Study of the Patterns and Structures of Wooden Floors
The manor house in Brzeziny was definitely created in accordance with a perfect
symmetrical plan, stressing its representative function. The manors could be more or
less sumptuous, depending on the ambitions or the social status of their owners, or on
the architect. The decorative parquets that have been preserved in the Witkowice manor
house, located nearby, include a representative rosette in the Banquet Room, evoking
the style of rich piano nobile parquets of the neighbouring Łańcut Castle. The parquets
preserved in two other rooms in Witkowice have a more practical character, although
some aesthetical ambitions can be seen in them as well (Fig. 3).
Fig. 3. Parquet patterns in manor houses in: Witkowice (a and b), Bieżdziedza (c) and Dydnia (d);
scale: 1:100 mm
The manor in Kolbuszowa is a typical wooden manor of late baroque, built in the
18th century, with a traditional layout of room functions (apart from the kitchen that
was introduced at the end of the 19th century). The stately rooms of the garden row
should have more sumptuous, decorative parquets. The side rooms of the front row
should have more humble decorative parquets. The last room of the garden row that has
been changed into a kitchen can receive a wooden “frieze” (with a pattern of squares
made of planks and battens) that will be both practical and decorative.
4 Summary
The patterns of high-class architecture passed to vernacular architecture with certain

delay. In existing manor houses, the shape of the building would be transformed and
improved, exterior walls would be covered with boarding and interior design would be
changed, covering the ceiling beams with a stylish lower ceiling (planks on the bottom
part of the beams), placing wooden panelling on the walls and changing parquet patterns.
Nonetheless, considering that manors had sentimental value for Polish nobility as the
seats of their families, their attachment to tradition and respect for the creations of their
ancestors limited the interventions in the shape of those buildings. On the other hand,
interiors were often modified, in order to follow the most fashionable trends in archi‐
tecture. In Polish manor houses since the beginning of the 19th century, panel parquets
and even decorative rosettes (e.g. Witkowice Manor) started to replace parquets made
of boards or boards organised in a pattern of squares separated with battens (the so called
frieze floors).
As a result of the introduction of collective property and the appropriation of manor
houses after World War II, the interior furnishing was irreversibly destroyed. At this
moment, only iconographic materials can show us the richness of many of those houses.
There is an urgent need of stock-taking and documenting the buildings that have been
preserved, as well as reconstructing the non-existing buildings together with their inte‐
rior furnishing. This problem is also very important due to the ongoing deterioration of
the interiors of historical buildings caused by continuous usage, changes of owners and
refurbishment works [4, 5].
References
1. Archer, B.L.: Systematic method of industrial design. In: Biblioteka Wzornictwa, Vol. 7.
Instytut Wzornictwa Przemysłowego, Warsaw (1987)
2. Fortuna-Marek, A., Gliwa, A., Potera, B.: Manor homestead as a functional layout and its
realisation in the Ethnographic Park of the Museum of Folk Culture in Kolbuszowa. Biuletyn
Muzeum Kultury Ludowej w Kolbuszowej 1, 113–141 (2009)
3. Sobolewski, H.: Detailed Design for a Manor Homestead - Brzeziny Manor House, Rzeszów
(2014)
4. Rozanska, A., Burawska, I., Policińska-Serwa, A., Koryciński, W., Mazurek, A., Beer, P.,
Swaczyna, I.: Study of antique wooden floor elements of chosen buildings from south-eastern
Poland. In: Proceeding of the 8th International Conference on Structural Analysis of Historical
Construction SAHC 2012, Wrocław, Poland, 15–17 October 2012
5. Rozanska, A., Sudol, E., Wierzbicki, J., Mazurek, A., Beer, P.: Antique wooden floor
construction solutions and the possibilities of using them in reconstructions. In: 2nd
International Conference on Structural Health Assessment of Timber Structures (SHATIS
2013), Trento, Italy, 4–6 September 2013
Project Paper: Heritage Building
Information Models (HBIM)
Digital Tools for Heritage Preservation and Enhancement
The Integration of Processes and Technologies on 20th Century
Buildings in Brazil and India
Luca Rossato ✉
( )
Department of Architecture, University of Ferrara, Ferrara, Italy

[email protected]
Abstract. Currently the 20th century architectures are all over the world in danger
and under attack: these buildings (in many cases designed by international
renowned professionals) are facing a silent destruction. Day by day they are
slowly modified in terms of materials, volumes, colours or even demolished.
These architectures are still used for public purposes or as residential buildings
but they are usually in bad conditions and their state of materials conservation is
quite poor. This on-going research explores in depth the possibility to preserve
and valorize modern heritage in Brazil and India by the integration of 3D tools,
processes and technologies in order to face the future preservation challenges.
Soon after the research process the main topics of the project were evaluated and
studied to create the bases for a structured research path. In parallel the case study
assessment was able to identify suitable buildings (both in Brazil and India) on
the which the chosen topics are now being applied in order to improve the knowl‐
edge on the design process and reach a guidelines proposal for the preservation
and enhancement of these architectures.
Keywords: 20th century architecture · Modern architecture · 3D tools ·

Technologies · Heritage enhancement
1 Subject of the Scientific Research
1.1 State of the Art

The global research on modern architecture started few decades ago but it has already
involved many historians, academics, curators, writers, archivist and journalist: people
who lent their knowledge and time in order to better understand a miscellaneous
phenomenon. In this wide research field the global exchange of architectural ideas and
forms over the 20th century is an important characteristic to be highlighted which came
from the geographical movements and migrations of architects and engineers.
Modern buildings are apparently less sensitive to minor changes as those of earlier
historic periods. To preserve this massive built heritage it is essential to understand their
history and fundamental design principles. If we really want to cure the buildings
diseases, like doctors do, we should know the anamnesis of patients, their background

DOI: 10.1007/978-3-319-48496-9_45
568 L. Rossato
and causes of illness. The design of windows, for example, often is the most interesting
part of their architectural characterization. Modern buildings are, for their concepts,
shapes, materials, perfect case studies to explore the integration of 3D technologies and
heritage conservation (and valorization toward the transmission of knowledge to future
generations).
The design of windows, for example, often is the most interesting part of their archi‐
tectural characterization1. Modern buildings are, for their concepts, shapes, materials,
perfect case studies to explore the integration of 3D technologies and heritage conser‐
vation concepts.
Despite the wide range of modernist buildings in global south countries (previously
known as developing countries), architecture history books focus on the Western
scenario of this style. As stated by Lu (2010, p. 1) “with the exception of the work of a
very small number of acclaimed architects, little attention was devoted to modern archi‐
tecture in third world which was considered merely lesser forms of Western
modernism”.
1.2 Anthropological Background

The research on the strong relation between the new modernism principle influence in
South America and Asia and the anthropological effect on local societies and styles could
help to better understand this tangible (and also intangible) heritage toward its preser‐
vation and enhancement. A more anthropological approach is needed to identify the
transcultural principles underlying the design of modern buildings in India and Brazil:
as recently stated by the Italian sociologist Domenico De Masi, these are currently the
last two countries bringing a sort of humanism within their deep cultures. On the one
hand Brazil is the land of the corporal humanism due to its image of “country of happi‐
ness”, where things are very often related to the curves of human body (for instance
Oscar Niemeyer architectures2), on the other hand India is the country of spiritual
humanism, a place where souls and thoughts merge and spread out a sense of deep
personal involvement in every visitor3.
Transformation/adaptation process on modernism buildings in Brazil and India
could be defined as a kind of architectural transculturation phenomenon. The term
transculturation has been coined by Cuban anthropologist Fernando Ortiz in 1947 to
describe the transition process of merging and converging cultures. Transculturation
does not consist merely of acquiring another culture (acculturation) or of losing or
uprooting a previous culture (deculturation). Rather, it merges these concepts and addi‐
tionally carries the idea of the consequent creation of new cultural phenomena (neocul‐
turation)4.
Brazil and India have strong relations with migration influxes or outflows that influ‐
enced local art and culture. Also the architectures of this countries have always reflected
1
Macdonald (1996).
2
Philippou (2008).
3
De Masi (2014).
4
Ortiz (2002).
Digital Tools for Heritage Preservation and Enhancement 569
this views, especially the ones from the twentieth century buildings designed by the
great masters of modernist architecture in a period of incredible flow of ideas and
concepts and sharing of cultural backgrounds.
In this wide scenario the nationalist movements coincide with the modernist archi‐
tectural thinking and the promotion of an international style. Architects responded to
the increasing challenges in different ways, some employed technological innovations,
others got inspired from vernacular solutions to design regional interpretations modern
architecture principles5.
2 Methodology
The research is exploring by deep study 3D technologies for the heritage conservation,
highlighting the current state of art, opportunities, processes and possible future chal‐
lenges.
The research themes are:
• Twentieth century architecture framework in India and Brazil
• Documentation processes and technologies
The Research process is based on the evaluation of:
• Architects migration flows and related transculturation phenomenon
• Materials of modern architectures in Brazil and India
• Preservation framework on modern architectures in Brazil and India
• Digital archives and database use for heritage enhancement
• 3D modelling for documentation
• Non-invasive technologies
In selection process the main topics of the research project were evaluated and
studied in order to create the bases for a structured research path. In parallel the case
study assessment identified suitable buildings as possible case studies. The main selec‐
tion process steps were:
• Selection of case studies (200 buildings analyzed)
• Selection of the most suitable topic to be applied on case studies
In application process the chosen topic are currently being applied to the most suit‐
able case studies in order to reach a design proposals guidelines based on local context
and environmental issues. The research main topics are:
• Data sheet and representation techniques (applied on 80 buildings out of
200)_Sect. 3.1
• BIM approach (applied on 20 buildings out of 200)_Sect. 3.2
• 3D laser scanner survey (applied on 3–4 buildings out of 200)_Sect. 3.3
5
Lang et al. (1997).
570 L. Rossato
The innovative building materials and structural systems of these modern architectures
are at the same time warts and all. On one hand modern architects were brilliant and
very creative on the other they not always had enough knowledge on the nature and
behavior of new materials.
This research could actually help towards the creation of local management plans
that could guide long-term maintenance of buildings and the transmission of modern
design principles (Fig. 1).
Fig. 1. The research methodology
3 An Heritage in Danger: Topics Application
3.1 Representation Techniques: The Redrawing Process of Modern Architecture

in India and Brazil
If time has slowly switched off the power of the transformer dreams of modernist archi‐
tects, their buildings are a legacy of extraordinary value that should be protected and
enhanced as it is a collection of real lessons of architecture. For this reason, the in-charge
teachers of the course of Techniques of Architectural Representation (TRA) a discipline
of the second year of Architecture Curriculum at the Department of Architecture of the
University of Ferrara, decided to analyze by different groups of students the heritage
built in the vast and heterogeneous Indian and Brazilian contexts.
In 2016, for the third consecutive year, through the works of the greatest architects
of the period it is possible to browse an important slice of history of architecture, passing
by the national identity search for specific local features, an architecture that is eclectic,
hybridized, which addresses the theme of living, of dwelling, with a completely new
and varied language of a different symbolism from that of the past, redesigned with
poetry and sharpness.
The use of drawings in order to carry out analytical reviews of the archival heritage
of 20th century architectures in India and Brazil can reveal design experiences that reflect
the different territorial contexts from which they emerged and the cultural forces behind
them. The so called “survey of the project” was the adopted methodology: by analyzing
and redrawing the original documents using innovative graphic layouts the research can
highlight the potentials of these built heritage.
The aspects taken into account in this phase describe the complexity of the study and
the need of well-structured data. The process led to a very good understanding of the
designer’s work by the interpretation of original drawings, scheme and pictures, that
reveal the design process behind the construction.
The deep knowledge of the buildings, carefully selected and analyzed with a precise
methodology and representation techniques, had as outputs an exceptional variety of
ideas for further research and reinterpretations. For example the study of demolished
modern buildings or the reconfiguration of different design hypothesis for the most
important buildings of this period (Fig. 2).
Fig. 2. Redrawing and understanding the modern architecture: Gandhi Ashram in Ahmedabad
by Charles correa. Research and drawings by A. Sousa, I. Bulgaru, M. Abdelhaj
3.2 BIM Approach: Modelling Ramkrishna House House in Ahmedabad, India,

by Charles Correa
The elaboration of digital models by Ferrara University Architecture Department staffs

has gone through a careful planning and guided process in the field of BIM (Building
Information Modeling), a virtual three-dimensional space in which each component can
be called full-scale, integrating all information related to the geometry with details
concerning the materials employed, the phases of realization, costs, technical charac‐
teristics, and by linking the building with environmental factors (Fig. 3).
572 L. Rossato
Fig. 3. Ramkrishna House in Ahmedabad, by Chalres Correa, Revit 2017® BIM model by
E. Ranuzzini, O. Ballardini, D. Felloni.
In this framework, the Development of geometrical single-disciplinary Building

Information Model (BIM) of the Ramkrishna House in Ahmedabad gave to research
team basic shared knowledge of the data harvested from the documentation analysis and
it will help to manage all information about the building.
The house, designed by Charles Correa, presents parallel walls which form the
backbone of its plan, a structure divided into 4 main zones: living areas, guest rooms
with private garden, service area and bedrooms on the upper floors.
Built between the 1960 and 1962 the house explores the ideas related to dwelling
issues in India in strong relation with climate challenges6.
The BIM model of this architecture has been created taking advantage of the docu‐
mentation research on Charles Correa drawings. The picture of the original construction
were thus utilised in order to study the materials framework. Beside being a strong base
spatial research and study by this model was also used in order to create a virtual reality
(VR) model. Software as V-Ray use a proprietary stereoscopic rendering camera to
generate a 3D environment with existing Revit cameras, lighting and materials. With
this environment it was so possible to generate photo-realistic images of the Correa
project to give to the public a sense of what the house look like by creating a VR “map”
of a building inside and out (Fig. 4).
The key passages of the BIM creation were focused on: acquisition of building data,
integration of other sources of information able to describe the built heritage, definition
of reference geometries for architectural components, development/detailing of the
6
Scriver and Srivastava (2015).
Fig. 4. Ramkrishna House in Ahmedabad, by Chalres Correa: original pictures (left side) and
photo-realistic views (right side). Rendering by E. Ranuzzini, O. Ballardini, D. Felloni.
building as 3D digital reconstruction, and implementation of instruments for the exploi‐

tation of the model between valorization and management.
The advantages associated to the approach proposed concern the flexibility of the
instruments adopted, the quality of the results obtained in terms of precision and level
of detail, and the possibility of communication between the product developed and other
solutions oriented on the exploitation of the building through the model.
The final output of this 3 years BIM research highlighted a strong synergy between
building management and valorization approaches through the workflow. Future
perspectives of this scenario are related to the diffusion of accurate methods for the
acquisition of building data and the development of 3D models. Building Information
Modelling methodology is increasingly penetrating into building design. Currently laser
scanning appears to be the ideal solution for data acquisition on existing buildings. It
allows the quick and precise high definition capture of 3D data. The challenge here is
to create a parametric 3D BIM model from the precise depiction of the real world, in
the form of a point cloud. For instance point snapping allows the precise remodelling
of the point cloud regions with 3D BIM elements.
This diffusion may bring a wide and common evolution, influencing management
strategies of cultural heritage currently adopted in India and Brazil with the optimization
of management strategies and improvement of knowledge on modern heritage.
3.3 3D Laser Scanner Integrated Survey of Casa das Canoas, by Oscar Niemeyer,
Rio de Janeiro, Brazil (1951)
Designed by Niemeyer in 1951 as his family home, Das Canoas House is considered
one of the most significant examples of modern architecture in Brazil and is well recog‐
nized by specialists in art history as a synthesis of modern architecture and authorial
self-creation that flourished in Europe and in America (Fig. 5).
574 L. Rossato
Fig. 5. Casa das Canoas (Rio de Janeiro) by Oscar Niemeyer
The most interesting feature is the fusion of organic architecture and minimalist
architecture. As it is an obvious tribute to the Farnsworth House, Das Canoas puts
geometric regularity in crisis, but does not alter the formal purity. From Mies design to
Niemeyer’s Das Canoas, its possible to see the evolution of reinforced concrete into a
modern building material7.
The house is currently facing bad problems of materials deterioration and lack of
maintenance. Main degradation issues are vegetation, that is growing very fast on
external walls and on horizontal surfaces due to the local climate; structural, because
part of the house is slowly moving towards the hill slope and an intervention need to be
urgently planned; big cracks cross the bedrooms ceiling and rain water is leaking inside
the house, plaster is becoming very fast darker and weaker; swimming pool pavement
shows cracks and deterioration and further analyses are required to better understand
the causes of this problem; great geological instability of the slope subject to strong
storms is a serious degradation factor.
The documentation of the architectural masterpiece by Niemeyer has been intended
priority to analyse the current condition and the state of conservation of the building,
which currently suffers from a lack of maintenance. The integrated survey was carried
Fig. 6. Casa das Canoas: points cloud section CAD ready for architectural drawings
7
Weinstraub and Hess (2012).
out by the DIAPReM centre at Ferrara University and started from the architectural
analysis: Oscar Niemeyer’s ideologies in terms of society, politics and architecture had
formed a unique style, which he expressed to the full in the architecture of the house at
Canoas, which in itself was a remarkable achievement of Modernist architecture (Fig. 6).
The documentation of the architectural masterpiece by Niemeyer has been intended
priority to analyse the current condition and the state of conservation of the building,
which currently suffers from a lack of maintenance. The integrated survey was carried
out by the DIAPReM centre at Ferrara University and started from the architectural
analysis: Oscar Niemeyer’s ideologies in terms of society, politics and architecture had
formed a unique style, which he expressed to the full in the architecture of the house at
Canoas, which in itself was a remarkable achievement of Modernist architecture.
The “reading” of the architecture composition focused three main themes of inves‐
tigation: the shape of the space and its three-dimensional use, the study of the relationship
between architecture and sculpture and the visual connections between interior and
exterior, the relationship of architecture with the surrounded green landscape (Fig. 7).
Fig. 7. Locations of main sculptures on point cloud plan and their relation with the Casa das
Canoas’ main openings
Niemeyer adopted this idea and throughout his career he not only designed sculptural
buildings, but he also insisted for the right sculptures and paintings to go with them. In
the Casa das Canoas are placed sculptures by his friend Alfredo Ceschiatti. For the first
time a 3D survey with this level of detail (3 mm) allows to identify new ways of “reading”
and cross interpretation, such as the triangulation perspective between the artwork
placed by Oscar Niemeyer into the enclosed and not enclosed space. Thanks to the
detailed data capturing it was possible to identify the axes of relationship that develop
from the placement and orientation of the female bodies to encourage to share the envi‐
ronmental and architectural experience.
576 L. Rossato
An integrated survey methodology was applied in order to obtain set of data to be

critically analysed to examine in depth the research themes on the architecture by
Niemeyer (Table 1).
Table 1. Main data of the instrumental survey campaign.

Total station Leica TS06
Matched targets 53
Polygon vertexes 7
3D laser scanner Leica C10
Number 3D scan stations 128
Acquired targets 142
Point cloud coordinates 1.717.862.695
Work schedule 10 days (about 75 h)
The integrated survey project involved 10 days on field working, about 8 h per day.
The total computed area for the 3D survey campaign included the external area of the
house and the garden. This is because the landscape project was an essential part of the
overall design process by Oscar Niemeyer. The 3D survey has been carried out with
Leica C10 Laser Scanner (Lidar technology) and topographic survey was performed by
Leica Total Station TS06 plus 2″. Topographical main connection to the network and
framework, consisting of a polygonal closed or opened to bench marks at the ends and
with polygonal rods with no longer than 80 m. The survey campaign has been docu‐
mented by a detailed photographic survey.
The great opportunity to document and survey one of the most significant architec‐
tures by Oscar Niemeyer, developed in cooperation with the Oscar Niemeyer Founda‐
tion, has shown how the integrated methodology is able to reveal new aspects of the
building and to analyze spaces and surfaces by means of innovative methods that have
allowed to track intellectual avenues completely unexplored and unpublished.
4 Future Steps
The series of documentation developed by the masters of the modernism in India and
Brazil that have been deeply analyzed during this research, have been reconfigured and
utilized in order to achieve the creation process of 3D study models. These models enable
greater and more effective understanding of the masterpieces of modern architecture in
these countries and thus a possible stronger preservation policy. Future steps should be
the creation of a digital platform by the which it will be possible for researchers and
experts to have access at this database.
On the other hand it has been highlighted that three-dimensional survey has proved
to be essential to represent buildings/areas that would be otherwise impossible to
analyze, for example elevations deeply immersed in very green areas, and to find plans
matches that are essential to understand the architectural “philosophy” of the modernism
masters.
In this representative phase there is a motivational value, which makes the survey-
representation a real project itself, with significant critical implications aimed at the
determination not only of geometric precision but, especially in architecture, of visual‐
ization and conceptual representation of reality.
At the moment integration of 3D survey and BIM approach seems to be the strongest
base for future conservation plans and valorization projects. Furthermore, in case of
demolition or unexpected events (such us natural disasters, etc.) this methodology will
also preserve the memory of this extraordinary heritage.
To address these challenges, the Getty Foundation in Los Angeles, USA, developed
Keeping It Modern, a grant initiative that aims at the conservation of modern buildings.
The programme is already supporting modern buildings grant projects of outstanding
architectural significance that promise to advance conservation practices: the University
of Ferrara is deeply involved in the preservation masterplan of three of this buildings.
The cooperation between the Getty initiative and the on-going research could actually
help towards the creation of conservation management plans that guide long-term main‐
tenance policies and the testing and analysis of modern architecture materials (Fig. 8).
Fig. 8. Some examples of 3D database of Indian and Brazilian modern architectures developed
by the research carried out at the University of Ferrara Architecture Department
References
Baan, I., Muller, L., Nooteboom, C., Stierli, M.: Brasilia-Chandigarh: Living with Modernity.
Lars Muller Publisher, Zurich (2010)
Bahga, S., Bahga, S.: Le Corbusier and Pierre Jeanneret: Footprints on the Sands of Indian
Architecture. Galgotia Pub. Co., New Delhi (2000)
Balzani, M., Sasso, D.F., Rossato, L.: The architectural survey for the revitalization of Villa Itororó
in São Paulo. Paesaggio Urbano-Urban Design 5–6, 68–75 (2015)
Belluardo, J., Ashraf, K.K.: An Architecture of Independence. Princeton Architectural Press, New
York (1999)
578 L. Rossato
Bhargava, M.L.: Architects of Indian Freedom Struggle. Deep & Deep, New Delhi (1981)
Bhatt, V., Scriver, P.: Contemporary Indian Architecture After the Masters. Mapin Publishing.
Pvt. Ltd, Ahmedabad (1990)
Brusaporci, S.: Emerging Digital Tools for Architectural Surveying Modelling and
Representation. Hershey, IGI Global (2015)
Cairo, A.: The Functional Art: An Introduction to Information Graphics and Visualization.
Pearson Education, Inc., New York (2012)
De Masi, D.: Mappa mundi: Modelli di vita per una società senza orientamento. Rizzoli, Milan
(2014)
Lang, J.T.: A Concise History of Modern Architecture in India. Permanent Black, Delhi (2002)
Lang, J.T., Desai, M., Desai, M.: Architecture and Independence: The Search for Identity - India
1880 to 1980. Oxford University Press, Oxford (1997)
Lu, D.: Third World Modernism: Architecture, Development and Identity. Taylor & Francis,
New York (2010)
Macdonald, S.: Modern Matters: Principles and Practice in Conserving Recent Architecture.
Donhead, London (1996)
Model House Research Group (ed.): Transcultural Modernisms. Sternberg Press. Wien (2014)
Montaner, J.M.: Arquitetura e crítica na América Latina. Romano Guerra Editora, São Paulo
(2014)
Ortiz, F.: Contrapunteo cubano del tabaco y el azúcar (Letras His). Catedra, Madrid (2002)
Philippou, S.: Oscar Niemeyer: curves of irreverence. Yale University Press, New Haven (2008)
Rossato, L.: The architectural survey and representation of the modern project between
preservation and sustainability. Paesaggio Urbano-Urban Design 4, 12–19 (2015)
Salonia, P.: Which is the right direction? Proceedings of the XVIII General Assembly ICOMOS,
Florence, 9 November 2015, pp. 521–523. Napoli (2015)
Scriver, P., Srivastava, A.: India: Modern Architectures in History. Reaktion Books, London
(2015)
Sowa, J.F.: Knowledge Representation: Logical, Philosophical and Computational Foundations.
Thomson Learing, Boston (2000)
Theoharis, D.: Contemporary Third World Architecture, Search for Identity. Pratt Manhattan
Center Gallery, New York (1983)
Waisman, M.: O interior da história. Perspectiva, São Paulo (2013)
Weinstraub, A., Hess, A.: Oscar Niemeyer Casas, GG Brasil, São Paulo (2012)
From Integrated Survey to the Parametric
Modeling of Degradations.
A Feasible Workflow
Massimiliano Lo Turco(&), Federico Caputo, and Gabriele Fusaro
Department of Architecture and Design, Politecnico di Torino, Turin, Italy

[email protected],
{federico.caputo,gabriele.fusaro}@studenti.polito.it
Abstract. This work fits into an international research field about 3D modeling
to evaluate the Building Information Model performance for infographic rep-
resentation of Cultural Heritage. Modeling an historic building involves the
creation of parametric objects library starting by data survey. The primary
purpose of the research is the translation of these information into a parametric
model, through the definition of a proper methodology. The main focus of the
research is the creation of parametric object representing the preservation status
of material and building components: some recurring schemes of the traditional
representation have been identified, in order to find a methodology that leads to
link these data to the HBIM (Historic BIM) model, improving their capabilities.
Keywords: Cultural heritage Point cloud H-BIM 3D decay Data

representation
1 Introduction
In the last period Building Information Modeling is emerging as the most reliable
method to manage architectural design and building processes. Literature supplies both
theoretical approaches and several practical applications. However, very little research
regards to BIM applied to historical architecture, even if some initial results indicate the
actual HBIM (Historic BIM) capability for the conservation of Cultural Heritage [1].
Since the architectural heritage has taken advantages from the development of data
capturing technologies, an open issue is the integration of BIM systems and data source
for existing buildings, supporting the creation of 3D knowledge and management
systems [2].
At this regard, after various heuristic processes, a novel workflow that starts from
point cloud and image survey data was defined: it was needed to define different steps:
1. Focus on the issue of accuracy between point cloud and modeled objects [3]
Contrary to what usually was done, the control on the level of geometrical accuracy
was directly done in the BIM environment, as shown below on Sect. 4.
2. Creation of historical libraries of building object that currently do not exist. Also in
this case, the current workflows do not provide for a direct upload of “cropped”

DOI: 10.1007/978-3-319-48496-9_46
580 M. Lo Turco et al.
point cloud into the design component environment. Several approaches were
tested; two of these were adopted (later explained on Sect. 4).
3. Finally, we focused on the chance to model 3D decays in the BIM platform,
enriching the related database with graphic, geometric and alphanumeric data that
can be effectively used to design and manage future interventions (Sect. 5). This
issue is considered by the authors as the most innovative part of the entire work,
because no instances related to these theme were found.
2 The State of the Art: A Brief Overview
According to the main statements of the London Charter for the computer-based
visualization of Cultural Heritage (that establishes internationally-recognized principles
for the use of computer-based visualization by researchers, educators and Cultural
Heritage organizations) it is necessary to take into account several types and degrees of
access that computer-based visualization can uniquely provide to cultural heritage
stakeholders, including the study of change over time, magnification, modification,
manipulation of virtual objects, embedding of datasets [4]. Basically, the accurate and
detailed reconstruction of geometric models of real objects has become a common
process. Most of the published works are mainly focused on data survey processing,
through a comparison between point clouds and models. Some others critically discuss
about the meaning of “rigorous” BIM protocols [5]. In literature, other works show
several steps in HBIM modeling, pursuing several aims, such as:
– energy evaluations and the searching for a new database for the construction
industry, as in InnovANCE project, which was also attended by some researchers of
the Politecnico di Torino [6].
– structural simulations and conservation analysis [7].
– monitoring processing, characterized by a strong contribution of Geomatics [8].
– More conceptual considerations on the rules of classical architecture [9, 10].
The proposed work relates to the validation of an effective system constituted by
graphical and alphanumeric representation; this usual BIM approach is implemented by
the infographic description of the materials and their preservation status that belong to a
historical heritage.
3 The Case Study of “Torino Smistamento” Roundhouse
The analyzed case study is a property of the Ferrovie dello Stato, now abandoned and
originally used for storage of locomotives, for which it has been proposed a specific
intervention. The building was created to adapt standard the railways for the Universal
Exhibition of 1911, to celebrate the 50th anniversary of the establishment of the
Unification of Italy. It is composed by two different areas: the first one is used for the
maintenance and the parking of the trains; the other one is used for the goods
switching, coming by railway. The maintenance area was constituted by a roundhouse
From Integrated Survey to the Parametric Modeling of Degradations 581
containing 51 shelted railroad tracks (Fig. 1), next to two sheds for the most complex
interventions (A and B parts) (Fig. 2).
Fig. 1. Detail of the site plan, developed in 1905, compared with the existing situation.
Development of the wares staging area of Torino Smistamento (nearby Lingotto) [11].
Fig. 2. The 1st December 1943 bombardment: birdseye view of bombings concentrated in via
Nizza [12].
The bombings of 1943 caused huge damage to the circular structure, knocking
down some of the 51 aisles, reducing the overall number to 32. The roundhouse is
made of a structural repeated module constituted by three bays, sized to host a
railroad-car. All rails are linked to a junction turning platform, in order to allocate the
wagons inside the module. Actually the building is partially used (Fig. 3).
Fig. 3. Perspective birdseye view of the roundhouse, taken from Google Earth, Turin,
acquisition date: 08/07/16, 45°02′28.66″N 7°40′00.37″E elev. 239 m alt. 353 m.
4 Description of the Survey and Parametric Modeling

Activities
After the historic research, the design survey has been realized; the work was char-
acterized by combined techniques, mostly based on TLS. As known, the Terrestrial
Laser Scanner can acquire a large amount of data in a very short time, automatically
recording geometric, radiometric and photographic information, later georeferenced
into a unique coordinate system.
The point cloud is described through statistical precision indicators that do not
actually define the true overall accuracy of the cloud. To do that, the reliability has been
verified through a test procedure which evaluates the differences between point coor-
dinates measured with the total station and point coordinates extracted from the cloud.
This process allowed to assess the overall tolerance, which is the parameter that really
matters to the user survey.
The point clouds are then imported into the BIM software to start with the modeling
phase. The production of an “As-Is” model involves the need to model the existing
architecture components, which do not belong to the objects contained into the usual
software libraries.
The creation of new elements implicates a reading of the geometric rules followed
to define the shape of the building components. It is therefore necessary to semantically
interpret the geometrical rules and then to parameterized them. As previously men-
tioned, a method for the creation of parametric objects starting from the point cloud
was experienced: that procedure is not feasible in the classic versions of BIM plat-
forms, because it is not allowed to import segmented point clouds within the interface
used to model single components.
Moreover, the main parameters of certain system families have also been cus-
tomized, such as the walls and slabs layering. At the same time, a geometric layout of
the building structure was defined: this was done replicating the module with a polar
array, setting the reference system of radial axes and determining the center of the
structure. These assumptions were validated after the comparison with the point cloud
(Fig. 4).
Fig. 4. Top view of the entire point cloud. In red, the distances between the points used for the
testing of measures. (Color figure online)
In order to achieve a better validation of the digital model, new parameters were
added: some for the GRADE (GRAphic DEtail that define the detail level of the
graphic representation) and LOD controls (Level of Development, meant as the degree
of reliability of information associated with different elements). These usually are used
in the design phase, when the information assets increase together with the design
development; some new variables were added to assess the reliability degree and
completeness of information element detected. Finally, this data can be viewed through
thematic representations.
The geometric validation of the model was performed experimenting the use of a
software for the metric control: at this regard, it was used Pointsense, a plugin for the
management of point clouds within Revit environment (BIM application used). This
plugin allows to evaluate the accuracy of the proposed virtual reconstruction, providing
some tools about flatness, verticality and horizontality controls, directly working with
BIM software. This control is explained in detail in most of the works previously
mentioned [2, 3, 5] (Fig. 5).
Fig. 5. The graphs provided by Pointsense allow to assess qualitatively the deviation between
the point cloud and parametric model: most of the points that belong to the surface of the walls
have a deviation that does not exceed 3 cm. The color changes on the wall surface are an index of
non-planarity. (Color figure online)
They actually use different software, managing the exported 3D model with a loss
of information; in other word the metric accuracy is evaluated by duplicating data with
a high risk of making mistakes: the process is not bidirectional, so it is not possible to
reimport this data into the BIM platform. But the proposed innovation identifies a more
interoperable workflow, as we proceed to calculate the difference between the virtual
position of an element and the data survey readable by the point cloud: if this difference
is less than the predetermined tolerance, the geometric reconstruction can be considered
reliable.
5 New Strategies for Representing Parametric Decays
The last part of testing involved research and critical analysis of the most innovative
representation of the internal decays. A useful method to define new ways of working
in BIM environment has been identified. The test focused on a sample form of the
structure: there, the preservation status has been detected, firstly plotted using the
traditional representation techniques. The methodology used refers to the handbook of
Mario Dalla Costa [13] which can still be considered a reference point for the
Restoration area. In this approach we have added some changes for entering
alphanumeric and graphic data into the parametric modeling environment. Regarding to
the usual representation method proposed in the manual, all the information on dating
and the consistency of the building are actually shown on the usual 2D drawings in
orthogonal projections, applying a pattern element connected to a legend containing
added data.
Within the BIM model, geometries are represented through well-defined mathe-
matical rules. The geometries of pathologies instead can be very complex. We have
tested several approaches (for instance: the implementation of the functionality through
programming languages such as API and Python, or a mixed use of algorithmic design
software, to name only the main ones). Then, it was decided to use an “adaptive
component” (for Revit is a particular object that can be adapted to a surface using
movable vertex) able to be associated with flat and curved surfaces (Fig. 6).
Fig. 6. On the top, left side: adaptive component based on 20 vertices. On the right, parametric
decays readable in 3D views. Below, the editing process and merge of the various parts.
Doing this, the vertex can be modified, and the decay components are visible in the
3D environment, supporting new ways of viewing and analysis. The added value
consists on the ability to associate new parameters that describe the pathology, the
working procedure to restore the building components, photos, data materials and
quantities (i.e. extensions of the decay, expressed in square meters).
So, multiple analysis can be done, making them more evident through multicriteria
schedules defined by qualitative and quantitative data, preserving the production of
traditional graphic drawing. At the end of this process a specific template has been set
up for future restoration works, containing all settings used for this pilot project.
6 Conclusions and Future Developments
This paper proposes a method able to describe, classify and organize information assets
concerned with Architectural Heritage, through the use of integrated survey
procedures [14].
The research constitutes a very little part of several studies already known by the
Scientific Community; however we are fully convinced of our innovative ideas for the
introduction of BIM methodology for documentation of architectural heritage, with a
specific focus on the restoration works (Fig. 7).
Fig. 7. On the top: isometric view of the modulus combined with the 3D representation of the
decays. Below, sample labels applied to a 3D decay, with the corresponding legend.
The work demonstrates that it is possible to develop a HBIM, starting from the
checking of the metric accuracy: in particular, the transfer from numerical model (cloud
of points) to a mathematical model (3D modeling) is a process that involves simpli-
fications and deductive hypothesis. Furthermore, these hypotheses depend both on the
object of the study and from the survey purposes: therefore these options must be
individually evaluated [15]. From the geometric point of view, the BIM platform
manages mathematical models, choosing the rule and not the form: introducing
Fig. 8. Thematic sheet of the current preservation state: section that shows the internal facade
with description of the degradation pathologies, realized through BIM application: any
information present on the sheet is directly associated to the 3D degradation element.
elements characterized by complex geometry, such as a pattern which describes decays,

is cumbersome and not so interesting. On the contrary, managing information related to
the preservation state can be considered perfectly compatible with BIM way of
thinking.
It is assumed to have made a research that aims at improving the efficiency and
effectiveness of more conventional procedures. The final model synthesizes the
information usually fragmented (firstly because belong to different representations)
enriching it with new methods of analysis and management of data, opening up the
possibility for new scenarios of “knowledge” [16].
Finally, we have argued about the replicability of the process, thought as the reuse
of objects that belongs to the architecture of the past, and for the creation of a working
template for future reusable surveys, to achieve the infographic sharing of models and
procedures (Fig. 8).
Acknowledgements. The authors would like to thank Kubit software for their kind authoriza-
tion to use Pointsense for Revit plug-in for research purposes. Thanks also to the Faro Europe,
(with specific regard to Alberto Sardo) for the availability, by providing the support needed to
implement the proposals simulations.
References
intelligence to laser and image based surveys of European classical architecture.
ISPRS J. Photogram. Remote Sens. 76, 89–102 (2013)
2. Quattrini, R., Malinverni, E.S., Clini, P., Nespeca, R., Orlietti, E.: From TLS to HBIM: high
quality semantically-aware 3D modeling of complex architecture. Int. Arch. Photogram.
Remote Sens. Spat. Inf. Sci. 40(5/W4), 367–374 (2015)
3. Biagini, C., Capone, P., Donato, V., Facchini N.: Towards the BIM implementation for
historical building restoration sites. Automation in Construction (2016)
4. http://www.londoncharter.org/
5. Barazzetti, L., Banfi, F., Brumana, R., Previtali, M.: Creation of parametric BIM objects
from point clouds using NURBS. Photogram. Rec. 30(152), 339–362 (2015)
6. Caffi, V., Daniotti, B., Lo Turco, M., Madeddu, D., Muscogiuri, M., Novello, G., Pavan, A.,
Pignataro, M.: Il processo edilizio supportato dal BIMM: l’approccio InnovANCE - BIMM
enabled construction processes: the InnovANCE approach. In: ISTeA, Bari, pp. 91–109,
10–11th July 2014
7. Dore, C., Murphy, M., McCarthy, S., Brechin, F., Casidy, C., Dirix, E.: Structural
simulations and conservation analysis-historic building information model (HBIM). Int.
Arch. Photogram. Remote Sens. Spat. Inf. Sci. 40(5/W4), 351–357 (2015)
8. Barazzetti, L., Banfi, F., Brumana, R., Previtali, M., Roncoroni, F.: BIM from laser scans…
not just for building: nurbs-based parametric modeling of a medieval bridge. ISPRS Annals
of the Photogrammetry and Remote Sensing, vol. III-5, XXIII ISPRS Congress, 12–19 July
2016, Prague, Czech Republic (2016)
9. Dore, C., Murphy, M.: Semi-automatic generation of as-built BIM facade geometry from
laser and image data. ITcon 19, 20–46 (2014)
10. De Luca, L.: Relevé et multi-représentations du patrimoine architectural Définition d’une

approche hybride de reconstruction d’édifices. Thèse de l’ecole doctorale Arts et Metiers
ParisTech, Aix-en-Provence, mars (2006)
11. Ballatore, L., Masi, F.: Torino, Porta Nuova: storia delle ferrovie piemontesi. Abete, Roma
(1988)
12. Bassignana, L.: Torino sotto le bombe. Nei rapporti inediti dell’aviazione alleata. Edizioni
del Capricorno, Torino (2003)
13. Dalla Costa, M.: Il progetto di restauro per la conservazione del costruito. Celid, Torino
(2000)
14. Inzerillo, L., Santagati, C.: Using dense stereo matching techniques in survey. Disegnare
Idee Immagini 47, 82–91 (2013)
15. Lo, Turco M., Sanna, M.: La modellazione digitale ricostruttiva: il caso della chiesa della
Misericordia a Torino. Disegnare Idee Immagini 41(2010), 42–51 (2010)
16. Docci, M., Maestri, D.: Manuale di rilevamento architettonico e urbano. Laterza, Bari (2009)
INCEPTION Standard for Heritage BIM Models
Peter Bonsma1 ✉ , Iveta Bonsma1, Anna Elisabetta Ziri2, Silvia Parenti2 ✉ ,

( ) ( )
Pedro Martín Lerones , José Luis Hernández , Federica Maietti , Marco Medici4,
3 3 4
Beatrice Turillazzi5, and Ernesto Iadanza5

1
RDF Ltd., Iskar 25, 1000 Sofia, Bulgaria
{peter.bonsma,iveta.bonsma}@rdf.bg
2
Nemoris srl, Via Decumana 74 A, 40133 Bologna, Italy
{annaelisabetta.ziri,silvia.parenti}@nemoris.it
3
Fundación CARTIF, Parque Tecnológico de Boecillo, P. 205,
47151 Boecillo, Valladolid, Spain
{pedler,josher}@cartif.es
4
Department of Architecture, University of Ferrara,
Via Ghiara 36, 44121 Ferrara, Italy
{federica.maietti,marco.medici}@unife.it
5
Consorzio Futuro in Ricerca,
Via Saragat 1, 44122 Ferrara, Italy
Abstract. The EU Project INCEPTION will create a platform that is able to

exchange content according to state-of-the-art available open BIM standards.
This INCEPTION open Heritage BIM platform is not only exchanging data
according to existing state-of-the-art standards, but it is based on a new Heritage
BIM model using Semantic Web technology. This allows applications to retrieve
content according to modern query languages like SPARQL and allows user
defined ‘on-the-fly’ extensions of the standard. This paper describes the structure
and development of this new Heritage BIM standard. The Heritage BIM standard
is developed by several Semantic Web and BIM standardization specialists in
combination with top experts in the field of Cultural Heritage, all of them partners
within the INCEPTION project.
Keywords: 3D · BIM · IFC · ifcOWL · GIS · H57 · H-BIM · OWL · OWL2 ·

Semantic web · RDF · RDFS
1 Introduction
The European Project “INCEPTION - Inclusive Cultural Heritage in Europe through

3D semantic modelling”1, funded by EC within the Programme Horizon 2020, focuses
1
The INCEPTION project, Grand Agreement no.: 665220 started the last June 2015, is devel‐
oped by a consortium of fourteen partners from ten European countries led by the Department
of Architecture of the University of Ferrara. More information can be found on [http://
www.inception-project.eu/].

DOI: 10.1007/978-3-319-48496-9_47
INCEPTION Standard for Heritage BIM Models 591
on three main objectives: to create an inclusive understanding of European cultural

identity and diversity by stimulating and facilitating collaborations across disciplines,
technologies and sectors; to develop cost-effective procedures and enhancements for on-
site 3D survey and reconstructions of cultural heritage buildings and sites; to develop
an open-standard Semantic Web platform for accessing, processing and sharing inter‐
operable digital models resulting from 3D survey and data capturing.
This inclusive approach includes open-standard format for cultural Heritage
Building Information Modelling (H-BIM) as part of the overall procedure aimed at
enriching and enhancing the changing role of 3D representations for knowledge, recon‐
struction, preservation and exploitation of Cultural Heritage.
The integration of semantic attributes with hierarchically and mutually aggregated
3D digital geometric models is essential for management of heritage information. The
development of tools for 3D automatic delineation depending on acquisition technolo‐
gies, from point clouds to photo-based data, allows to achieve a common standard inter‐
operable output for BIM environment. Therefore, starting from advanced procedures
aimed at handling multi-data point clouds and triangle meshes into BIM software, the
INCEPTION procedure advances BIM approach for Heritage knowledge, going a step
forward the usual procedure to locate/define 2D or 3D primitive shapes onto the point
clouds. INCEPTION develops methodologies and algorithms to recognize these shapes.
Results will be constructed in BIM software avoiding the oversimplification of the
shapes. When used in models of Cultural Heritage, semantic BIM will be able to be
connected to different users (e.g. scholars, technicians, citizens, etc.) in support of the
user’s needs for interpretation of the cultural heritage model, in addition to the common
BIM features of 3D visualization, technical specifications and dataset.
The recent earthquake in central Italy (23 August 2016), causing about 300 victims,
almost destroyed the beautiful towns of Accumoli and Amatrice. The last one was
inserted in 2015 among the “most beautiful villages of Italy”. Both the towns date back
to XI–XII centuries and are home to beautiful churches and sanctuaries, heavily
damaged by the earthquake. The same fate has befallen many houses of historical
interest. Similarly, in 2009, the same region was hit by a big earthquake that fatally
wounded the wonderful ancient city of L’Aquila, still undergoing a slow process of
architectural recovery, causing 309 victims.
One of the aims of the INCEPTION procedure is protecting the cultural heritage of
seismic areas with scopes of classification, prevention and reconstruction.
The paper starts exploring the state of the art within existing open standards, focusing
on available H-BIM solutions up to explaining INCEPTION implementations.
2 Existing Open Standards
In the area of BIM, GIS, Cultural Heritage and Semantic Web, a lot of valuable work is
already done. INCEPTION has taken existing state-of-the-art open standards and tech‐
nology as a starting point. In this chapter we will just name a few relevant open standards
in the area of BIM and Point Clouds, without being complete in number of standards
592 P. Bonsma et al.
nor in the areas (for example GIS was removed completely keeping the paper size
reasonable). A complete reference can be found in D4.1 from the INCEPTION project.
2.1 Existing BIM Standards

This paragraph will explain the main open BIM standards expected to be used for data
providers of BIM.
IFC (ifcXML, ifcOWL, ifczip). The first developments for the IFC format dates back
to 1985. The name IFC (Industry Foundation Classes) was first introduced in 1994 led
by Autodesk. In 1995 it became a vendor independent standard and had several releases,
IFC 151 and IFC 20LF (Long Form) were popular releases for the academic world. Since
IFC 2 × 3 released in 2006 (and later improvements IFC 2 × 3 Final and IFC 2 × 3 TC1)
it was also becoming more popular for use in real life projects. Nowadays IFC 2 × 3
TC1 is still the most popular version although for INCEPTION the new version IFC4
(formally known as IFC 2 × 4) has some important improvements for both 3D repre‐
sentation and scheduling.
IFC is meant to be used by all the disciplines in the Building & Construction industry
and the only widely used open standard supporting so many different disciplines. All
major CAD vendors and a wide variety of other applications offer support for IFC. Most
of the applications supporting IFC are not certified, although most CAD systems with
IFC support have a certification from buildingSMART. IFC carries an object-based view
of the model, including geometry in 3D (and 2D) as well as properties and interrela‐
tionships between objects. It is a well thought through standard with relatively high
complexity for software vendors to support. IFC support includes schedule data, quan‐
tities and many other construction related data.
The standard exchange format for IFC is STEP/EXPRESS. As serialization support
for this format is limited the past few years other serializations are defined also. It started
with support for ifcXML, with an alternative ‘simple’ ifcXML format. More recently
also ifcOWL serialization is created, this last format is compatible with Semantic Web
RDF, RDFS and OWL2 standards. Although there is a small data loss the fast majority
of knowledge is kept in these alternative serializations. A different format is ifczip and
is nothing more than the zipped version of an original IFC file.
bSDD. Building Smart Data Dictionary is like the semantic extension of the IFC
schema. Although IFC in combination with its property sets (and about 3000 properties)
has already a lot of semantics in it, to cover the complete Building & Construction
industry the semantic definition has to be far larger (numbers differ but we could expect
that 100.000 object definitions are required where even the latest IFC schema has less
than 1000 entities). bsDD is the standard from Building Smart defining how such exten‐
sions of the semantics can be stored and defined.
2.2 Existing Point-Cloud Standards

E57. Most 3D imaging systems for data exchange today takes place using one of three
types of file formats: proprietary formats (not an efficient approach to data exchange in
the long term), ad-hoc formats (not space or time efficient and no widespread usage), or
the LAS format (limited file size and features). The E57 format is intended to overcome
these issues, being a more general format that is well-suited for storing data across a
variety of application domains. It is able to store point clouds and also other information
from 3D scanners like images. The file format is specified by the ASTM, an international
standards organization, and it is documented in the ASTM E2807 standard Huber (2011).
Next to the standard a ‘reference implementation’ is created to make more attractive
and easy the use of the standard. The reference implementation is called libE57 and is
written in C++ and sources are available [http://www.libe57.org/]. The libE57 appli‐
cation contains an API that can be used by parties that like to import or export files in
E57 format.
2.3 Available H-BIM Solutions
Several H-BIM Solutions are already available. One thing we can notice in many of
these solutions is use of Semantic Web techniques or use of BIM related standards. For
example the vendor specific standard Graphisoft GDL language (Graphical Description
Language) allows parametric modelling of components, something very useful in the
area of geometry for Cultural Heritage content.
One typical behavior of most currently available H-BIM solutions is that they are
clearly focused on one or two areas of the core of a Semantic Web based solution:
• Cultural Heritage
• Semantic Web technology
• BIM /3D /Point Cloud knowledge
3 Semantic Web
3.1 RDF
RDF (Resource Description Framework) supports creating and processing metadata by
defining a default structure. This structure can be used for any data, independent of their
character. Thus, the application areas of RDF are numerous, e.g., web-based services,
peer-to-peer networks, and semantic caching models; they all have in common that huge
amounts of data have to be processed when querying RDF data. RDF data can be repre‐
sented using XML, a triple structure or a graph. Only the graph representation enables
the semantic interpretation of the RDF schema.
All of the elements of the triple are resources with the exception of the last element,
object, that can be also a literal. Literal, in the RDF sense, is a constant string value such
as string or number. Literals can be either plain literals (without type) or typed literals
typed using XML Datatypes. These triples together form RDF graph. A normative
syntax for serializing RDF is RDF/XML.
3.2 RDF Schema (RDFS)

RDFS extends RDF vocabulary to allow describing taxonomies of classes and proper‐
ties. It also extends definitions for some of the elements of RDF; for example it sets the
domain and range of properties and relates the RDF classes and properties into taxon‐
omies using the RDFS vocabulary.
The RDF schema statements are valid RDF statements because their structure
follows the structure of the RDF data model. The only difference to a pure “resource -
property - value” - triple is that an agreement about the specific meaning for reserved
terms and statements has been made. Next to that, the RDF schema provides a vocabu‐
lary for defining the semantics of RDF statements.
3.3 Web Ontology Language (OWL)
OWL is a W3C standard. The abbreviation stands for Web Ontology Language and is
a language for processing information on the web. It is built on top of RDF and RDFS.
OWL was designed to be interpreted by computers and parsed by applications. It is not
meant for being read by people. OWL is written in XML and has three sublanguages -
OWL Lite, OWL DL (includes OWL Lite) and OWL Full (includes OWL DL). The
Ontology is about the exact description of things and their relationships. For the web,
ontology is about the exact description of web information and relationships between
web information. The standard OWL is a part of the “Semantic Web Vision”, a future
web where:
• Web information has exact meaning
• Web information can be processed by computers
• Computers can integrate information from the web
3.4 Web Ontology Language 2 (OWL2)
OWL 2 adds new functionalities with respect to OWL 1. Some of the new features are
syntactic sugar (e.g., disjoint union of classes) while others offer new expressivity,
including keys, property chains, richer data types, data ranges, qualified cardinality
restrictions, asymmetric, reflexive, and disjoint properties, and enhanced annotation
capabilities. OWL 2 also defines three new profiles and a new syntax. Some of the
restrictions applicable to OWL DL have been relaxed resulting in a slightly larger set
of RDF Graphs that can be handled by Description Logics reasoners.
3.5 Reasoning
OWL enables “reasoning”, as mentioned above. That means it gives the possibility to
check the logical correctness of statements and add statements that are implied by other
statements
A “semantic reasoner”, “reasoning engine”, “rules engine”, or simply a “reasoner”,
is a piece of software able to infer logical consequences from a set of asserted facts or
axioms. The notion of semantic reasoner generalizes that concept of inference engine,
by providing a richer set of mechanisms to work with. The inference rules are commonly
specified by means of an ontology language, and often a description language. Many
reasoners use first-order predicate logic to perform reasoning; inference commonly
proceeds by forward chaining and backward chaining. There are also examples of prob‐
abilistic reasoners, including Pei Wang’s non-axiomatic reasoning system, and Nova‐
mente’s probabilistic logic network.
3.6 SPARQL
SPARQL is the reasoning language for Semantic Web. Servers support in many cases
out-of-the-box SPARQL queries.
The gate for the access to the Inception ontology will be a SPARQL endpoint.
SPARQL 1.1 is a semantic query language and a recommendation of W3C. Its adoption
in Inception project is important to access to the CH semantic storage. Multiple program‐
ming languages, libraries and semantic repositories implement SPARQL queries. The
1.1 standard also allows to write queries which directly update the RDF graph. Being
SPARQL syntax based on graph traversal, it is also easy to visualize SPARQL results
graphically. Although repositories could be navigated and examined with different tools,
a SPARQL endpoint is one of the powerful tools to open semantic data to main exploi‐
tation.
4 H-BIM Ontology
This chapter describes the architecture of the H-BIM Ontology as well as some examples
of the content. The H-BIM Ontology is the core of the INCEPTION Platform. The
content of the Ontology will be developed together with the specialists that are partners
within INCEPTION. By definition the content will however be incomplete; the archi‐
tecture therefore allows users to extend the H-BIM Ontology either for projects or for
larger aggregations, like countries, styles etc.
4.1 Background
The architecture of the H-BIM Ontology is defined to enable storage of semantic infor‐
mation from any cultural heritage object. A clear distinction between tangible and non-
tangible content is defined and, in line with the base concepts behind Semantic Web
technology, it allows layered extension of the ontology itself. Typical for the H-BIM
Ontology is the close connection with existing state-of-the-art BIM standards like IFC/
ifcOWL and the link with 3D content defined in Semantic Web (i.e. open standard CMO
with Extensions).
4.2 Architecture Overview
4.3 Technology Layer
Within this layer we use the Semantic Web technology as described in Sect. 3. This
means use of the RDF, RDFS and OWL2 as top layers of the H-BIM Ontology.
4.4 INCEPTION Layer

The INCEPTION layer of the H-BIM model contains the real knowledge from the
specialists within the INCEPTION project. This is knowledge about Cultural Heritage
but also knowledge from existing state-of-the-art open BIM and open GIS standards.
One important part of the INCEPTION layer is the difference between tangible and
non-tangible results. Since this is a known term in the area of Cultural Heritage and
distinction is not always clear, much time and effort is put in defining what is covered
by the terms.
4.5 Extensions Layer

As soon as new concepts within the extension layer are defined the queries can be used.
Extension could concern new relations between existing content and therefore applied
on all existing content, but extensions can also be specific for dedicated content. Some
examples:
Example of Extensions Applicable on Existing Content. The INCEPTION H-BIM

standard has embedded parts of the semantic structure of IFC (and therefore ifcOWL).
This means classes Wall, WallStandardCase and Cur-tainWall exist. Adding a super
class Walls and the knowledge that above named classes inherit from this new class
Wall is an extension that works on all available content. A SPARQL query can be created
to get all instances of new class Walls and it will directly have content for the majority
of the Cultural Heritage H-BIM models stored in the INCEPTION platform.
Example of Extensions Applicable on New Content. It is allowed to add a class to

the H-BIM model in the extension layer with a new name and no relation to any existing
part of the H-BIM model. A query on this class is only relevant for new content incor‐
porating the knowledge that this new class is existing.
4.6 Project Layer
Within the project layer the real content is defined, this content is arranged according to
the layers above. All content can be queried according to the SPARQL queries defined
on top of the INCEPTION layer. It is also possible to create solution specific queries as
well as queries dedicated to certain extensions as defined in the extension layer.
5 Implementation
The INCEPTION standard is the base for the platform that will be developed within
INCEPTION. As the INCEPTION standard is developed.
5.1 Server Solutions
The core of the INCEPTION platform will be a server that is able to handle the INCEP‐
TION standard and offer basic functionality like support for SPARQL. During writing
of this paper the selection of the server handling this Semantic Web data is not finalized
yet, there are several options and the most promising solutions at this moment seem to
be RDF4 J and Fuseki 2.
As not only Semantic Web data needs to be stored, but many different file formats
including open standard BIM formats as well as point cloud data a file server will be
part of the INCEPTION platform also.
Sesame /RDF4J 2.0. The official name is RDF4J and its current release is 2.0, however
this solution is best known under its former name OpenRDF Sesame framework. It
became part of the Eclipse Foundation and has no official first version at this moment
after this important change.
Fuseki 2. Fuseki server is already existing for a while and although especially Fuseki
2 would be of interest for INCEPTION also Fuseki 1 is still actively maintaine. Fuseiki
2 is a server solution on top of Apache and Jena and also called Apache Jena Fuseki. It
is a SPARQL server and an open source project.
5.2 SPARQL Queries
Technology choices and implementation will be driven by the accessibility of data

through SPARQL queries addressed to a SPARQL endpoint. This means that a running
REST web service should respond to data queries and produce as results a set of triples
serialized on one of the most used RDF serialization (RDF/XML, Turtle, N3) or an RDF
graph. This should include queries both on tangible and non-tangible data and their
specific relations.
SPARQL queries are “data-oriented” so there is no inference in the query language
itself, all the data manipulation and inferencing has to be done by a layer on storage or
on in memory RDF data. Thanks to its structure and many converters present in the
market SPARQL queries can be applied not only to native RDF data but also on any
data that could be mapped to RDF, like other kind of well-formed relational data.
6 Conclusion
The use of BIM for Cultural Heritage is becoming more and more an effective tool to
manage 3D representations at different layers and for multiple purposes, pursuing the
common vision, at European level, to apply research, technology and innovation in
innovative media to expand understanding and access of the heritage assets.
One of the main challenges is how to manage the complexity of heritage buildings
and sites, fostering the collaboration across disciplines through semantic-aware repre‐
sentations, able to solve interoperability issues and avoiding the segmentation of knowl‐
edge. The technology of Semantic Web and integration with 3D and BIM are the drivers
behind H-BIM Ontology. The H-BIM Ontology in its turn is the core of the INCEPTION
H-BIM Platform.
The INCEPTION procedure could be conveniently exploited for protecting the
cultural heritage of seismic areas with scopes of classification, prevention and recon‐
struction.
References
Bonsma, P.: Semantic web platform and interfaces. In: Deliverable 6.2 of the EU Project Proficient
(2014)
Huber, D.: The ASTM E57 file format for 3D imaging data exchange. In: Proceedings of the SPIE,
Electronics Imaging Science and Technology Conference (IS&T), 3D Imaging Metrology,
vol. 7864A, January 2011
Dore, C., Murphy, M.: Integration of Historic Building Information Modeling (HBIM) and 3D
GIS for recording and managing cultural heritage sites. In: 2012 18th International Conference
on Virtual Systems and Multimedia (VSMM), pp. 369–376. IEEE (2012)
Hichri, N., Stefani, C., De Luca, L.; Veron, P.; Hamon, G.: From point cloud to BIM: a survey of
existing approaches. In: International Archives of the Photogrammetry, Remote Sensing and
Spatial Information Sciences, vol. XL-5/W2 (2013)
Volk, R., Stengel, J., Schultmann, F.: Building Information Modeling (BIM) for existing
buildings - literature review and future needs. Autom. Constr. 38, 109–127 (2014)
Kauppinen, T., Paakkarinen, P., Mäkelä, E., Kuittinen, H., Väätäinen, J., Hyvönen, E.: Geospatio-
temporal Semantic Web for Cultural Heritage (2010)
BauDenkMalNetz – Creating a Semantically Annotated Web Resource of Historical Buildings-
Anca Dumitrache and Christoph Lange
Mantegari, G., Palmonari, M., Vizzari, G.: Rapid prototyping a semantic web application for
cultural heritage: the case of MANTIC. In: Aroyo, L., Antoniou, G., Hyvönen, E., Teije, A.,
Stuckenschmidt, H., Cabral, L., Tudorache, T. (eds.) ESWC 2010. LNCS, vol. 6089, pp. 406–
Pauwels, P., Verstraeten, R., De Meyer, R., Van Campenhout, J.: Architectural Information
Modelling in Construction History (2009)
From SfM to Semantic-Aware BIM Objects
of Architectural Elements
Massimiliano Lo Turco1 and Cettina Santagati2 ✉

( )
1
Department of Architecture and Design, Politecnico di Torino, 10125 Turin, Italy
[email protected]
2
Department of Civil Engineering and Architecture, Università di Catania, Catania, Italy
[email protected]
Abstract. The huge diffusion of Building Information Modeling approaches in

the field of architectural design has characterized the research of the last decades;
however very little research has been undertaken to explore the advantages and
criticalities of BIM methodologies in Cultural Heritage domain. Moreover, the
last developments in digital photogrammetry lead to easily generate reliable low
cost 3D textured models, that can be used to create semantic-aware objects of
reusable library of historical architectural elements. The aim is to test a novel
workflow practitioner centered, based on the use of the latest solutions for point
cloud managing into BIM.
Keywords: Cultural heritage · Point cloud · H-BIM · Structure From Motion

(SfM) · Data representation
1 Introduction
The use of digital technologies has totally changed and improved the working methods
applied to architectural heritage domain. Nevertheless, most operators, trusting in the
progressive automation of data acquisition, believe they are be able to manage the
complexity of reality without having the required cultural background. Only after a
scholar’s interpretation, data turn into an evaluated and recognizable information,
distinctive for the knowledge of the studied object.
The capabilities of Historical Building Information Modeling (H-BIM) need to be
explored, giving centrality to the critical interpretation of the acquired data: this is a
field of research always more essential and critical, also considering the copious pres‐
ence of historical buildings in Europe and in Italy and the lack of BIM protocols and
procedures on this topic. BIM approach can be defined as a “shared digital representation
of physical and functional characteristics of any built object […] which forms a reliable
basis for decisions” [1]. However, the suitability of BIM platforms for historical archi‐
tecture is still considered a great challenge. The first relevant issue is the relationship
between 3D data acquisition and their conversion in parametric semantic-aware compo‐
nents, hierarchically organized.

DOI: 10.1007/978-3-319-48496-9_48
From SfM to Semantic-Aware BIM Objects of Architectural Elements 601
The latest 3D acquisition techniques (laser scanning, digital photogrammetry)

supports easy and quick data acquisition. The obtained point cloud contains geometrical,
material and color data of the acquired objects, as well as information related to the state
of conservation. Nevertheless, the point cloud is constituted by a huge number of points,
ontologically indistinct until they are geometrically interpreted [2]. To this day, there is
no automatic processing tool available that allows for the identification of complex
shapes (such as the ones that characterize historical buildings) and turns them into
geometric parametric models.
It is a fact that the use of a commercial BIM package is mandatory to fulfill basic
BIM requirements [3]. However, the existing BIM platforms were developed for archi‐
tectural design; this means that the parametric objects inside them, often are not suitable
for the modeling of existing historical architecture elements such as walls, portals,
windows, cornices and string courses that use the elements of architectural language.
Furthermore, dealing with a historical building, in addition to geometrical informa‐
tion, a lot of heterogeneous data have to be organized and managed. The BIM database is
able to recover such information, becoming an indispensable resource for different profes‐
sionals involved in the restoration, intervention or management of existing buildings.
Therefore, it is necessary to create a shared parametric semantic-aware library of
architectural elements that belong to the different historical ages [4, 5]. These objects
have to be created as external components (families according to Revit). If well modeled
and generalized, these components can be reused in similar contexts. One chance is to
use the architectural treatises that describe the architectural classical elements, another
one is to start from the survey of architectural elements.
We apply this last approach, reasoning for single architectural components, devel‐
oping and testing a workflow that exploits the advantages of low cost digital photo‐
grammetry and of a specific plugin able to manage point clouds into commercial BIM
platform, thus useful to create parametric semantic 3D models of an architectural
element. We will reflect on data conversion (level of abstraction and generalization,
level of accuracy), Grade and LoD requirements in the case of “as-built” H-BIM. The
aim is to discuss these open issues, testing a novel workflow based on the use of the
latest available solutions for point cloud managing.
Here the structure of the paper: Sect. 2 will mark the state of the art and will analyze
different approaches and results; Sect. 3 describes the proposed workflow in detail; while
Sect. 4 refers to a chosen case study; then we will critically discuss the obtained results,
taking into account some feasible future research.
2 The State of the Art
In the last years, the use of BIM in the field architectural design field has brought to a
consolidation in the procedures and the identification of standard methods. However,
the challenge is still open if considering the conservation, management and enhancement
of the architectural heritage. It is required to review and update data acquisition process,
the standardization and structuring of acquired data in a 3D semantic model and the
subsequent representation and usability of the model.
602 M. Lo Turco and C. Santagati
Together, the digital recording of cultural heritage sites using laser scanning and
photogrammetry has become a topic of great interest in the field of conservation and
cultural heritage. Although data collection technologies are now very efficient and auto‐
mated, the processing of this data is still time consuming [6].
The spread of Image-based 3D modeling techniques, through digital photogram‐
metry free, low cost and open source packages, have drastically increased in the past
few years, especially in the field of Cultural Heritage [7]. Many researchers have carried
out several tests that have demonstrated the metric reliability of Structure from Motion
(SfM) techniques for architectural elements where other techniques (such as Terrestrial
Laser Scanning- TLS) are costly, not sufficiently dense or not easy to access [8, 9].
Moreover, tests performed by several research teams have demonstrated that these tech‐
niques are strictly dependent on the quality of the dataset (network, image resolution,
radiometric quality). [10] and are suitable for medium size objects, such as architectural
elements and details.
At present, we can distinguish between desktop and web based packages. If the first
ones need a high performance computer for data processing, the second ones use the
power of cloud computing to carry out a full-automatic data processing. In both cases,
the output is a dense textured point cloud of the analyzed object that can be easily turned
into a 3D mesh model.
Regarding to H-BIM methodology, very little research has been undertaken to
understand the potential of BIM for heritage buildings [11]. The efforts of researchers
are currently focused on two basic questions:
• Can be effectively used a BIM-based approach for the investigation on historical
buildings using commercial BIM platforms?
• How turn point clouds into rigorous BIM?
The first question deals with the lack of specific components/tools for historical
architecture available for commercial BIM platforms. The reconstruction of complex
shapes seems a challenging task. Once having obtained the point cloud and identified
the single elements and their mutual relationship, the operator could [12]:
– built an in-place family directly in the project environment;
– create a family that could be reused in other projects (usually BIM platforms don’t
allow to import point clouds into family editor except using specific plugins [13]);
– create 3D external objects and import them in the BIM model as surface models.
In literature, other works show several steps in 3D HBIM modeling [5]; these workflows
use different software with the necessary format conversion and we mainly observe a
2D simplification with slices of point clouds to build up the 3D model.
The second question is linked to the meaning of “rigorous BIM”. In literature we
find several studies [3–5] that address the crucial transition from the point cloud to the
intelligent parametric object, introducing the concept of level of accuracy. In other
words, the point cloud can be considered a digital copy of the object that preserves its
geometric features (irregularities, deformations, etc.): are we able to guarantee the metric
accuracy captured by digital survey in the BIM modeling phase? Is the level of abstrac‐
tion too high for an appropriate geometry reconstruction? Some authors carry out a
comparison - point cloud to model - to evaluate if the deviation is in keeping with the
scope of the H-BIM. Others prefer to perform the 3D modeling in other platforms able
to create and manage NURBS surfaces [3, 5]: they use proper protocols to convert
NURBS into parametric surfaces then uploaded in BIM platforms. Other studies refer
to “rigorous BIM”, meaning as the complete exploitation of BIM approaches for Cultural
Heritage buildings not only in terms of geometric accuracy, but also considering other
features (parametric objects, relations, attributes, correct setting of Grade and LoD).
2.1 Process and Product Taxonomies

Some samples of reconstruction of the existing condition often suggest to generate a
semantic model [14] based on a constant comparison between the information included
in the historical treatises and the profiles achievable from a point clouds. Some others
are focused on the issues of accuracy between building objects models (walls, pillars,
vault) and point clouds. Others, instead, suggest creating a historical library of building
object models currently not available.
Moreover, others propose a different classification of the whole approach, shifting
the focus from general to specific issues. In this regard, “as-built” BIM characterization
involves three aspects, allowing building a structured point cloud: shapes, relations and
attributes [15]. Regarding the shape of the object it can be classified according to three
variable: parametric or non-parametric/global or local/explicit or implicit [16]. In a
global representation, the entire object is described, while in a local one only a portion
of the object is characterized.
Regarding to the taxonomies of each single building component, a further important
consideration must be pointed out before developing this topic: as said, 3D models that
contain only 3D (metric) data are not BIM models [17]. This means that a model gener‐
ated from point clouds is not a BIM unless are added: parametric intelligence/relation‐
ships/attributes. Therefore, our challenge is to set up new digital components that belong
to cultural heritage, working on parametric through explicit modeling (CSG), and
preparing the field for corresponding data population.
3 A Workflow Centered on Data Critical Interpretation

for the Creation of Parametric Semantic Components
As highlighted in the previous works, the transition from the discrete (point cloud) to
the continuous (parametric 3D model created on the photogrammetric surveys imported
into BIM platform) still retain a certain stiffness. The first problem lies in the limited
usability of parametric components available in libraries and websites, because they
provide large amounts of material mainly focused on new constructions.
The evaluation of BIM capabilities applied to Cultural Heritage is a relevant field of
investigation, through heuristic approaches that could open new research avenues. Our
workflow starts from the acquisition phase: the point cloud survey requires a series of
pre and post processing stages, which involve cleaning, sorting and combining of
different sets of point cloud data. Then, point cloud data can be considered as a skeletal
framework, which is mapped using parametric elements to shape the H-BIM component
[5]. It must be said that BIM software mostly used, (Revit 2016), does not allow you to
import point clouds portions, in the family design interface, to be used as the basis for
the virtual reconstruction of the building components (doors, windows, pillars, beams,
…), the only vector formats that can be imported are .dwg or .dwf. The idea to create
“in-place family” has been voluntarily discarded, because such elements, directly real‐
ized into the model, can only be saved in the .rvt project (Revit file format), without
generating any .rfa external (and reusable) family. This last step is mandatory to set up
any library of reusable objects for subsequent interventions.
An elementary solution, although very time-consuming, consists on the use of appli‐
cations that handle point clouds to make a controlled selection, a subsequent conversion
into a text file and a further file processing to obtain the file format requested. The first
attempt considered a step for converting the point cloud into a .dwg file. This involved
the coding of a Python script able to permit the importation in Rhinoceros several exten‐
sions point file (.xyz, .pts, .csv) and the subsequent saving into .dwg file. However, once
imported into family design interface, the .dwg point cloud has some stiffnesses, making
the virtual modeling of certain items that needs to be visualized, analyzed and measured
in lateral views (sections) extremely arduous.
Then, we moved on commercial plugins that have been released in the last years
(such as Pointsense and Cloudworks for Revit). These tools make the management and
processing of point cloud easier in Revit family interface for example by extracting
orthophotos or segmenting point clouds. In this research work we tested the use of
PointSense for Revit, a Kubit plugin resell by Faro [21] that allows the user to import
the whole point cloud or a part of it directly into Revit family editor (Fig. 1).
Fig. 1. From right to left: RGB point cloud available on Revit interface; the point cloud loses the
RGB property when imported into Revit family interface; Pointsense xray ortho-view.
Furthermore, this plugin allows you to easily extract x-ray ortho-view from the point-
cloud which are georeferenced in the same reference system. So, ortho-images and
segmented/cropped point cloud are the basic data imported for further processing within
the family editor. Doing that, the critic interpretation of data is favored because point
data redundancy is abstracted to the essential object skeleton lines. This approach is
similar to the use of in situ eidotype as trace, with the advantage to have a point cloud
as reference that can be visualized, cropped and consulted in realtime. This brings to a
proper data interpretation and discretization: through the use of reference planes and
reference lines the skeleton of the new parametric component can be outlined; the
following dimension and conversion into parametric variables will guarantee the
geometric flexibility of BIM components.
Furthermore, each single element, if modeled in three dimensions, is able to store a
plurality of heterogeneous type information. This can be considered the added value of
using BIM instead of more established conventional approaches.
Model components and decorations were modeled through family design interface:
for each windows, doors and all the decoration and ornaments, we cropped the point
cloud to extract 2D profile useful to outline complex shapes drawn from classical orders.
According to Murphy [5] and De Luca [19], a bottom-up approach is adopted which
starts with the smallest building objects such as ornamental mouldings and profiles.
These uniform objects are created from a shape vocabulary of 2D shapes usable for all
configurations of the classical orders [20]. This approach is considered the best one,
because usable in future interventions, as it allows to identify the regulatory geometric
rules (invariants) (Fig. 2).
Fig. 2. Cima parametrization according to Aubin (A), De Luca (B), and family editor (C).
After the modeling phase we worked on the database to better organize the different
information gathered. All documents used for modeling, in fact, were linked to their
virtualized components, associating an image parameter to the different elements cate‐
gories. This implementation allows to link together in a unique virtual environment the
various original sources. The virtual reconstruction permits multiple queries and the
production of thematic drawings (such as the identification of the different measurement
instruments, the evaluation of the types of degradation, the fourth dimension control,
etc…).
4 A First Case Study Application
To test our workflow, we choose a 15th century gothic portal, characteristic of Catalan-
Aragonese architecture in the Etnean area of Eastern Sicily as case study. This portal is
one of the few memories that survived the catastrophic events at the end of the 17th
century in Eastern and South-eastern Sicily: the disruptive Mount Etna eruption (1669)
that covered and erased 16 etnean towns and the earthquake (1693) that destroyed almost
all the towns of the Val di Noto. The portal belongs to the old church of S. Maria delle
Grazie in Misterbianco (5 km far from Catania) which was covered by the eruption of
1669 and was brought to light in the last years thanks to the excavations carried out by
the Superintendence to Cultural Heritage of Catania. This portal represents a recurrent
typology of the catalan-aragonese architecture in Etnean area (Catania), other very
similar examples can be found in Mascalucia (church of S. Annunziata in Mompilieri
and the church of S. Antonio Abate) as well as in the remaining area (Randazzo, Fran‐
cavilla, S. Lucia del Mela, Taormina, etc.) (Fig. 3).
Fig. 3. 15th century Catalan-Aragonese gothic portals (from left to right): S. Maria delle Grazie
in Misterbianco; S. Annunziata in Mompilieri; S. Antonio Abate in Mascalucia.
The latest finding of the portal in Misterbianco, validates the hypothesis of a well
consolidate school of local craftsmen who used the available materials, interpreting and
elaborating the spanish style of the time. As a matter of fact, the presence of the lava
stone as construction material together with white limestone brings to a distinctive bi-
chromatism typical of these areas: the portal in Misterbianco is a round arch portal whose
jambs are in lava stone; the springing cornice ashlar is realized in white limestone as
well as the archivolt, the cordon and the conclusive corbel.
4.1 3D Survey
The photos were taken with a SONY DSC-W310 digital camera, 35 mm lens and at a
resolution of 12 Mpix. The shot project took in consideration the geometrical features
of the portal and the presence of the decorations for a total of 30 shots.
In order to design a workflow compliant with the most used commercial BIM plat‐
form (Revit) among all the available SfM packages we choose Autodesk cloud-based
photogrammetry service Recap 360. Taking advantage of the photogrammetric
approach and algorithms of Computer Vision, the service reconstructs the internal
parameters of the camera and the position in space of homologous points between frames
starting from the correspondence between a sequence of images. The user creates the
project, adds survey points or reference distances to scale the model, chooses the reso‐
lution of the model (low, ultra), the smart cropping and/or texturing option, the export
format (e.g. rcs, rcm, obj). Then the images are uploaded and sent to the cloud. When
the model is ready, the user can improve the results by adding survey points and resub‐
mitting the project. The obtained point cloud in rcs format was imported in Recap PRO,
then cropped and prepared for the following importation in Revit. Together, the obj
model (mesh textured model) was opened in Rhinoceros and the first textured ortho‐
photos were extracted (Fig. 4).
Fig. 4. View of the dataset, the reconstructed point cloud and orthophoto of the portal.
4.2 Geometrical Interpretation of Architectural Elements
The point cloud provides a digital replica of the portal, a mass of raw data that describes
the object but needs to be interpreted in order to be transformed into a parametric
semantic-aware model. To do this, the geometrical rules that generate the complex
surfaces have to be recognized as well as the corresponding mathematical law. The
construction of the geometric model requires a deep knowledge of the shapes and the
language of historical architecture. In this phase the scholar works consciously, recog‐
nizing the different elements and proceeding from general to particular by identifying
the architectural elements that characterize the object of study, gradually destructuring
the elements that constitute them up to the mouldings, the atoms of the architectural
lexicon [19].
Therefore, it is mandatory the rigorous identification of: the elements that geomet‐
rically describe the investigated surfaces; the geometric genesis of studied surface
(translation, rotation, interpolation); the mathematical law (Revolution, Estrusion); the
construction planes for each 2D profile and path and their extrapolation from the point
cloud by means of cutting planes. The architectural elements need to be decomposed in
their formal, material and structural features, to extract the geometrical rules that
generate the several surfaces/elements (invariant features) and a set of 2D profiles
(variant features).
In this sample, the portal presents some relevant features of Catalan Architecture: a
simple round-arch archivolt ended by a cordon and corbels (in limestone); straight jams
(in lava stone), and an ashlar that holds the arch spring cornice (in limestone).
Thinking on the invariant and variant elements, we can assume that not always is
the arch spring cornice present (as in the case of Mompilieri), and in some cases the
cordon profile is inside the external jamb line (once again in Mompilieri); furthermore,
the cornice and cordon 2D profiles varies according to the creativity of the craftsmen.
Therefore, portal invariant features are: the presence of a limestone archivolt ended
by a cordon (extrusion along a curve path), the presence of lava stone jambs (extrusion);
the presence of a corbel, the use of limestone ashlar in correspondence of arch springer;
the variant elements are the 2D profiles of the cordon, the presence of arch springer
cornice, the position of the 2D profile of the cordon (inside or outside the external jamb
line). All these issues have to be considered during the modeling phase, choosing prop‐
erly the reference planes, the origin of 2D profiles, the generating rules (extrusion) in
order to create a H-BIM component that generalize this type of Catalan-Aragonese portal
so that it can be re-used in other projects (Fig. 5).
The last step is the verification of the so called level of accuracy, that is the adherence
of the modeled component to the point cloud irregularities at the required scale of repre‐
sentation. This test can be directly carried out into BIM environment by using Pointsense
plugin or external software such as Geomagic, Cloud Compare, Meshlab.
Fig. 5. Modeling of the portal through the interpretation of the data contained in the point cloud
and list of type and instance parameters associated with the modeled family.
4.3 Focus on Grade and LoD Specification

As known, the BIM platform allows to associate multiple data to a single virtual compo‐
nent. The first group controls graphic contents: international guidelines identify the
Grade level to manage the graphical representation of building components in orthog‐
onal projection and spatial views, congruent with different levels of detail.
The second group concerns with alphanumeric data: the international standard iden‐
tify the LoD parameter (meant as Level of Development) as the degree of reliability of
information that can be expected from data contained into the digital model.
For what concerns the graphical views associated to different levels of detail, we
based the modeling procedure on the basis of eidotypes drawn during the survey, consis‐
tent to the different representation scales (Fig. 6); it was decided to set up:
– at Grade 1: displaying the jambs and the frame of the archivolt, highlight the different
material of springer cornice and jambs (consistent to 1:200 scale);
– at Grade 2: 2D schematic drawings of corbel and cordon were added (1:100 scale);
– at Grade 3: 3D modeling of the corbels, cordon; the ashlars of the archivolt and jambs
were more defined (1:50 scale).
Referring to LoD specification, it must be said that such classifications are closely
useful for new buildings interventions, where the measurement of the LoD level is linked
to economic, topological, construction and maintenance information. In the case of
interventions on historical buildings, it is needed to include other variables, critically
analyzing the richness of information available; this is done to measure the reliability
degree of the survey. More the survey is complete, greater the integration with the
various stakeholders who participated in the study can be arranged.
Fig. 6. Proposed Grade levels associated to the portal family.
This procedure includes the retrieval of design archives, the photos of the state-of-
art, the metric survey techniques, the degradation surveys. Through the creation of
shared parameters (usable to multiple types of components, and on several projects) we
are able to associate new data to the element detected. Then, the added information will
therefore be made evident in the model through labels, schedules or thematic views
(Fig. 7).
Fig. 7. 3D Family and list of shared parameters relating to survey procedures used to tag elements.
On the right, the corbel family uploaded on the portal and used as a nested family.
4.4 Results and Future Works

We tested our workflow on a single case study to fully apply our procedure, verifying
both theoretical and practical outlined issues. The promising obtained results lead us to
plan for the future. We will expand this work comparing our workflow applied to the
same object using several software and workflows from SfM object (point cloud) impor‐
tation to BIM, critically evaluating the results.
Moreover, we will test our workflow on more objects characterized by different kind
of complexity using the same approach, comparing the results and taking into account
the advantages and weakness.
5 Conclusions and Open Items
Regarding to existing H-BIM library objects, both the methods for semi-automatically
plotting building facades and manual plotting methods can be used. The approach used
with H-BIM is to map the objects in 2D onto segmented point clouds and orthographic
images in elevation, plan and section.
The sub-elements (mouldings, profiles, symbols, etc.) become the architectural
vocabulary; the whole composition relates to a linguistic structure, this linguistic
analogy offers architecture a basis for analysis and understanding. As a matter of fact,
ontology can be intended as a particular conceptual framework or as a specification of
a conceptualization. It enables aggregation, topological and directional relationships.
Aggregation (i.e. part of, belong to, etc.), could be modeled with a hierarchical-based
tree representation that permits to describe the composition in a local-to-global way.
Accordingly, it is possible to develop semantics and management procedures in order
to determine the correct LoD of the surveying and of the model. This can be surely
considered an open research topic, because of a current lacking in regulation systems
(or guidelines) that define the levels of LoD according to the degree of reliability of the
survey. The creation of a geometric model allows for many more applications of survey
data, such as semantic and information modelling which enables complex analysis,
management and visualization of heritage data.
Finally, it is crucial to make a general thought on the methodological accuracy: at
this regard, the London Charter [21], defines the principles to be followed for the 3D
representation of the Cultural Heritage, in line with the values of transparency, commu‐
nicability of the methods and results of this modeling process. So, this guarantees a
repeatability of the scientific process where the variable element is the data, the fixed
one is the process [22]. The application of these principles will allow to define a meth‐
odology for the knowledge (and the representation) of the Cultural Heritage that makes
the virtual reconstruction, the processing and communication of data more transparent.
The understanding of cultural dimension is thus a greater formal qualification in a
permanent relationship between architectural space and information space. It therefore
guarantees a repeatability of the scientific process where the variable element is the data,
the invariant one is the process.
Acknowledgements. The authors would like to thank Kubit software and Faro Europe (Alberto
Sardo) for their kind authorization to use Pointsense for Revit plug-in for research purposes.
References
1. ISO Standard, ISO 29481-1:2010(E): Building Information Modeling — Information

Delivery Manual — Part 1: Methodology and Format (2010)
2. Santagati, C.: 3D laser scanner aimed to architectural heritage survey: from the point’s cloud
to the geometrical genesis determination. Int. Arch Photogramm. Remote Sens. Spat. Inform.
Sci. 36 (5) (2005)
3. Barazzetti, L., Banfi, F., Brumana, R., Previtali, M.: Creation of parametric BIM objects from
point clouds using NURBS. Photogram. Record. 30(152), 339–362 (2015)
4. Quattrini, R., Malinverni, E.S., Clini, P., Nespeca, R., Orlietti, E.: From TLS to HBIM: high
quality semantically-aware 3D modeling of complex architecture. Int. Arch. Photogramm.
Remote Sens. Spat. Inform. Sci. 40(5/W4), 367–374 (2015)
intelligence to laser and image based surveys of European classical architecture. ISPRS J.
Photogram. Remote Sens. 76, 89–102 (2013)
6. Dore, C., Murphy, M.: Semi-automatic generation of as-built BIM facade geometry from laser
and image data. ITcon 19, 20–46 (2014)
7. Remondino, F., El-Hakim, S.: Image-based 3D modelling: a review. Photogram. Rec. 21(115),
269–291 (2006)
8. Santagati, C., Inzerillo, L., Di Paola, F.: Image-based modeling techniques for architectural
heritage 3d digitalization: limits and potentialities. Int. Arch. Photogramm. Remote Sens.
Spat. Inform. Sci. XL-5/W2, 550–560 (2013)
9. Kersten, T., Lindstaedt, M.: Automatic 3D object reconstruction from multiple images for
architectural, cultural heritage and archaeological applications using open-source software
and web services. Photogrammetrie - Fernerkundung - Geoinformation 2012(6), 727–740
(2012)
10. Gaiani, M., Remondino, F., Apollonio, F.I., Ballabeni, A.: An advanced pre-processing
pipeline to improve automated photogrammetric reconstructions of architectural scenes. J.
Remote Sens. 8(3), 178 (2016)
11. Fai, S., Graham, K., Duckworth, T., Wood, N., Attar, R.: Building information modelling and
heritage documentation. In: 23rd CIPA Symposium, Prague, Czech Republic, 8 p. (2011)
12. Biagini, C., Capone, P., Donato, V., Facchini N.: Towards the BIM implementation for
historical building restoration sites. Automation in Construction (2016)
13. Garagnani, S., Manferdini, A.M.: Parametric accuracy: building information modeling
process applied to the cultural heritage preservation. Int. Arch. Photogramm. Remote Sens.
Spat. Inform. Sci. XL-5/W1, 87–92 (2013)
14. Apollonio, F.I., Gaiani, M., Sun, Z.: 3D modeling and data enrichment in digital reconstruction
of architectural heritage. Int. Arch. Photogramm. Remote Sens. Spat. Inform. Sci.
XL-5/W2, 43–48 (2013)
15. Hichri, N., Stefani, C., De Luca, L., Veron, P.: Review of the “as-built BIM” approaches. Int.
Arch. Photogramm. Remote Sens. Spat. Inform. Sci. 40(5/W1), 107–112 (2013)
16. Tang, P., Huber, D., Akinci, B., Lipman, R., Lytle, A.: Automatic reconstruction of as-built
building information models from laser-scanned point clouds: a review of related techniques.
Autom. Constr. 19, 829–843 (2010)
17. Arayci, Y.: Towards building information modelling for existing structures. Struct. Surv.
26(3), 210–222 (2008)
18. http://faro-3d-software.com/CAD/Products/PointSense/3D_Laser_Scanner_Revit.php
19. De Luca, L.: Relevé et multi-représentations du patrimoine architectural Définition d’une
approche hybride de reconstruction d’édifices. Thèse de l’ecole doctorale Arts et Metiers
ParisTech, Aix-en-Provence, mars (2006)
20. Aubin, P.: Renaissance Revit: Creating Classical Architecture With Modern Software. G3B
Press, Oak Lown (2013)
21. http://www.londoncharter.org/
22. Novello, G., Lo Turco, M.: Which drawing to deliver more information? In: Less More
Architecture Design Landscape, Le vie dei Mercanti X Forum Internazionale di Studi, Aversa-
Capri (ITA) Napoli: La scuola di Pitagora editrice, pp. 690–699 (2012)
Project Paper: Novel Approaches to
Landscapes in Cultural Heritage
Observing Landscape Changes Around
the Nicosia Old Town Center
Using Multi-temporal Datasets
Branka Cuca(&), Athos Agapiou, and Diofantos G. Hadjimitsis
Department of Civil Engineering and Geomatics,

Cyprus University of Technology, Saripolou 2-8, Limassol, Cyprus
{branka.cuca,athos.agapiou,d.hadjimitsis}@cut.ac.cy
Abstract. In 1980s a significant boom in construction industry was witnessed

in Cyprus. This paper explores the changes of land use that have occurred over
the past 30 years around the historical capital of Nicosia, in particular around the
core of the historic city defined by the Venetian walls. Further to some Open
geospatial Data available within the national and regional geo-portals, the
research has focused on the use and exploitation of freely accessible satellite
imagery (such as Landsat and Sentinel imagery) and other archive aerial datasets
in order to observe the most recent modifications of the urban landscapes. The
changes occurred over time were observed using multi-spectral multi-temporal
dataset with main aim to create thematic maps for further interpretation. The
changes were hence identified, mapped and structured so as to emphasise dif-
ferent types and density of urban development affecting the surrounding land-
scapes and potential “hot-spots”. Such observations could be a valuable input to
the future urban development of Nicosia.
Keywords: Historical cities Nicosia Cyprus Earth observation Landsat

Changing landscape
1 Introduction
Geo-spatial information is increasingly being used for purposes of territorial man-

agement as it has been recognized to be crucial when it comes to informed decision
making processes that regard landscapes and environment. The nature of
geo-information is in fact twofold - on one side the elaboration of such data is capable
to describe phenomena that has already occurred (i.e. impact assessment purposes) and
on the other side to provide us with simulations of possible further effects or occurrence
(scenario modelling). In this paper authors opted for a combined and integrated use of
multi-spectral, multi-temporal and multi-source data available over Cyprus in order to
explore the contribution of freely available satellite imagery for monitoring of changes
around the city of Nicosia and to suggest possibilities offered by such data for
implementing the Open geospatial information available through national and EU
repositories. Open geospatial Data approach was also of interest due to the recently
established national INSPIRE platform, implemented by the Department of Land and
Surveying of the Republic of Cyprus [1]. In the framework of Open Government
DOI: 10.1007/978-3-319-48496-9_49
616 B. Cuca et al.
initiative, experts consider geospatial data to be significantly important category [2].

The possibility to access publicly collected and available data is retained crucial for the
development of new and innovative applications useful for a variety of social domains
[3]. Such approach that promotes the use and re-use of publicly collected data is
actually encouraged by public policies for economic purposes [4].
This paper investigates the contribution of satellite imagery to non-space sectors
such as built environment and urban development. The interaction between urban
(built) environment and the surrounding landscapes was observed applying a combined
and integrated use of multi-spectral and multi-temporal data (Sect. 2). Section 3
illustrates the methodology applied on the satellite remote sensing data with focus on
change detection, with illustration of the results in Sect. 4. A discussion on the method
and its use for purposes of decision making in urban contexts is provided in Sect. 5.
Paper concludes with inputs on the land use changes around the city of Nicosia and on
possible contribution of the information deriving from the satellite imagery in the future
decision-making processes that regard the island’s capital.
2 Datasets
The main data used to observe the changes in landscape around Nicosia’s walls were
freely available satellite imagery Landsat (LT), described in Table 1.
Table 1. Landsat imagery used in the study

Year Satellite (Sensor) Spatial resolution (m) Spectral resolution
(date-month) Panchromatic VIS (nm) (only VIS-VNIR
listed)
1987 LT5 (Thematic 15 30 450–900
(22-April) Mapper – TM)
2003 LT5 (Thematic 15 30 450–900
(25-Sept) Mapper – TM)
2016 Landsat8 Operational 15 30 433–885
(5-April) Land Imager (OLI)
The images were accessed and downloaded from the geo-portal of the Geographical
Survey of United States (USGS) for the years 1987, 2003 and 2016. This choice was
made in order to investigate changes in the period of 80s and 90s (years of construction
sector “boom” recorded in Cyprus) [5] and to observe possible modifications in urban
fabric in the years after Cyprus accession to the European Union in 2004. PCA
methodology was applied as it is frequently used to understand interactions between
landscape and urban sprawl. In 2015, for example, Agapiou et al. [5], use PCA to
determine areas of Paphos municipality that have undergone dramatic change in their
land use for the period 1984–2010.
Observing Landscape Changes Around the Nicosia Old Town Center 617
Figure 1 above shows Nicosia observed in all three images in Red – Green – Blue
(RGB) bands i.e. the band combination that corresponds to the human vision. Below
the same area as seen in band combination NIR – R – G that enhances the presence of
vegetation (in red). This paper concentrates on the changes occurred in the part of
Nicosia under administration of Republic of Cyprus, south of the buffer zone area.
Fig. 1. Nicosia Municipality seen in RGB (above) and as NIR – green – blue (below). (Color
figure online)
3 Methodology
3.1 Creation of Multi-spectral Multi-temporal Imagery Using Landsat
Data
In order to observe the land changes using this multi-spectral imagery over a period of
time, first it was necessary to build “stack layers” composed of two images. This was
performed for three pairs (1984–2003; 1984–2016 and 2003–2016) using ERDAS
Imagine 2010 software©. The images obtained in such a way are hence referred to as
multi-spectral multi-temporal imagery. However, it is important to mention that a
selection of bands for every image was made, according to their suitability for
observing changes of construction and vegetation areas i.e. the urban development and
possible landscape changes around inhabited spaces. Table 2 reports these combina-
tions for all three pairs used.
For experiment purposes an additional image combining all three images was built
following the same procedure. In this case, a composite image of 19 bands was
obtained, as illustrated in Table 3.
Figure 2 shows two scenarios A and B with hatched buffer zone (left) and with
buffer zone show as outline (right). The couple of images A refer to a stack image pair
1987–2016 observed with a combination 1,8, 14 i.e. showing the blue bands of both
618 B. Cuca et al.
Table 2. Landsat satellite imagery and respective bands used for creation of multi-spectral
multi-temporal pairs.
Image pair Single satellite Band number in the original Band number in
image per year image (wavelength) multi-temporal image
Pair 1987–2003 LT5 1987 Band 1 – blue (0.45–0.52) 1
Band 2 – green (0.52–0.60) 2
Band 3 – red (0.63–0.69) 3
Band 4 – Near Infrared (0.77– 4
0.90)
Band 5 – short-wave Infrared 5
(1.55–1.75)
Band 6 – Thermal Infrared 6
(10.40–12.50)
Band 7 – red (2.09–2.35) 7
LT5 2003 Band 1 – blue (0.45–0.52) 8
Band 2 – green (0.52–0.60) 9
Band 3 – red (0.63–0.69) 10
Band 4 – Near Infrared NIR 11
(0.77–0.90)
Band 5 – short-wave 12
Infrared SWIR (1.55–1.75)
Pair 1987–2016 LT5 1987 Band 1 – blue (0.45–0.52) 1
Band 2 – green (0.52–0.60) 2
Band 3 – red (0.63–0.69) 3
0.90)
(1.55–1.75)
Band 6 – Thermal Infrared 6
(10.40-12.50)
Band 7 – red (2.09–2.35) 7
LDCM 8 2016 Band 2 – blue (0.45–0.51) 8
Band 3 – green (0.53–0.59) 9
Band 4 – red (0.64–0.67) 10
Band 5 – NIR (0.85–0.88) 11
(1.57-1.65)
Band 7 – SWIR (2.11–2.29) 13
Band 8 Panchromatic (0.50– 14
0.68)
Pair 2003–2016 LT 5 2003 Band 1 – blue (0.45–0.52) 1
Band 2 – green (0.52–0.60) 2
Band 3 – red (0.63–0.69) 3
(continued)
Table 2. (continued)
Image pair Single satellite Band number in the original Band number in
image per year image (wavelength) multi-temporal image
(0.77–0.90)
LDCM8 2016 Band 2 – blue (0.45–0.51) 6
Band 3 – green (0.53–0.59) 7
Band 4 – red (0.64–0.67) 8
Band 5 – NIR (0.85–0.88) 9
(1.57–1.65)
Band 7 – SWIR (2.11–2.29) 11
Band 8 Panchromatic (0.50– 12
0.68)
Table 3. Landsat satellite imagery and respective bands used for creation of a composite
multi-spectral multi-temporal image.
Composite image Single satellite image Band at the origin and its Band number in the
per year wavelength domain multi-temporal image
1987 – 2003 – 2016 LT5 1987 Band 1 – blue (0.45–0.52) 1
Band 2 – green (0.52–0.60) 2
Band 3 – red (0.63–0.69) 3
0.90)
(1.55–1.75)
Band 6 – Thermal Infrared (10.40– 6
12.50)
Band 7 – red (2.09–2.35) 7
LT5 2003 Band 1 – blue (0.45–0.52) 8
Band 2 – green (0.52–0.60) 9
Band 3 – red (0.63–0.69) 10
(0.77–0.90)
LDCM8 2016 Band 2 – blue (0.45–0.51) 13
Band 3 – green (0.53–0.59) 14
Band 4 – red (0.64–0.67) 15
Band 5 - NIR (0.85–0.88) 16
(1.57–1.65)
Band 7 – SWIR (2.11–2.29) 18
Band 8 Panchromatic (0.50–0.68) 19
620 B. Cuca et al.
Fig. 2. Blue band combinations: (A) 1987–2016 image observed in band combination 1, 8 and
14 (above) and (B) 1987-2003-2016 image observed in band combination 1, 8, 13 (below).
images and using panchromatic band to enhance the spatial resolution. Couple B refers
to a stack of all three images (Table 2) observed with a band combination 1, 8, 13 i.e.
the blue bands in all three years. Such visualization criteria was chosen because
man-made structures (and materials such as cement) are more easily distinguished in
this range of spectrum.
3.2 3.1 Principal Component Analysis of Multi-spectral Multi-temporal

Data
In order to analyze these first observations, a spectral enhancement technique was
performed using a Principal Component Analysis (PCA). The parameter of 5 elements
was used in order to have a significant qualitative gradient of changes in a range from 1
to 5 (with 1 referring to the most and 5 referring to least significant changes, respec-
tively). Equivalent analysis was made for a multi-temporal image combining all three
datasets. The image obtained however results in a complicated set of information that
regards almost 30 years. In order to comment such image a more thorough background
in terms of historical information and documentation is needed (Fig. 3).
4 Results
The results obtained using the PCA is illustrated in Fig. 4. It is to be noted that such
images appear in false-colour meaning that RGB channels are attributed to single
elements of the PCA. Here the band combination used was 1, 2, 4 meaning that higher
changes are observed in channels of Red and Green (1 and 2 respectively) while milder
changes are observed in Blue channel (4 on the scale of 1–5).
Hence, the colours seen on the image do not have an absolute value but are an
indication of landscape change in the cover (and hence use) observed and it requires
further qualitative and quantitative interpretation. A closer look on the walled city of
Nicosia and its southern area within the PCA image 1987–2016 (Fig. 5) show some
significant changes around the city walls (white dashed lined), just out of the walls and
along the main transport arteries of the city (red colour).
Such observations were identified as ‘hot-spots’ and were further examined
observing the values of sample pixels across all the bands used in both images.
Fig. 3. Principal Component Analysis for image pairs: 1987–2003 (left); 2003–2016 (middle)
and 1987–2016 (right). Buffer zone is shown as hatched (above) and as outline only (below).
622 B. Cuca et al.
Fig. 4. PCA of the image pair 1987–2016, a close-up on southern Nicosia. (Color figure online)
Using 1987–2016 PCA image as a reference, a sample of 9 points was selected

(points within the walled city, on the walls and out of the walls on ‘hot-spots’ and on
areas with medium and low change-rate). Choosing the blue band as the most suitable
one for observations of changes in construction, the absolute differences in reflectance
values were observed in bands 1 and 8 of the image pair 1987–2016 (Table 2), shown
in Fig. 5. The difference value is not steady but shows fluctuation: such behavior could
indicate changes in the urban environment (such as demolition or new construction) for
example in point 3 (around Venetian walls) or in point 6 (hot-spot 3). Further, the four
pixel samples corresponding to ‘hot-spots’ were examined for their reflectance values
Fig. 5. The absolute difference in blue bands in 1987 and in 2016 for all 9 sample points
Fig. 6. The absolute difference in NIR bands in 1987 and in 2016 for all ‘hot-spots’
in Near Infrared (NIR) band in both years (i.e. bands 4 and 11 from Table 2). Figure 6
shows that differences in reflectance of ‘hot-spots’ 1, 2 and 3 have a similar trend, while
it is different for ‘hot-spot’ 4. NIR range of the spectrum this band is particularly
suitable for detecting changes in vegetation. This observation could hence be an
indication of change in vegetation cover in the ‘hot-spot 4’.
5 Discussion
The results illustrated show that quite a few areas in Nicosia have been subject to land
cover change and hence probably their land use. The PCA seems to be an adequate
method for qualitative analysis and useful for identification of ‘hot-spots’ i.e. areas
mostly subject to change. The limits of this approach is that areas cannot be assumed as
‘hot-spots’ only upon the investigation of satellite imagery but they need to be further
examined using urban masterplans or technical maps. In addition to reflectance pixel
values, the assumptions made on the type of land cover change are to be further
elaborated using historic data and in-situ measurements, also due to the limit of the
image spatial resolution. The method proposed can be however suitable for two rea-
sons: (1) it can provide an overview of the changes on a territorial scale of entire
municipalities (or even larger areas) and (2) it can help to identify specific areas of the
city that have been subject to change and that require further investigation.
6 Conclusions
The study here illustrated was an attempt to observe changes in urban environment of
the Nicosia municipality, in particular in the southern area around the walled historic
city. The methodology employed regards the use of freely available satellite imagery
that was examined using multi-spectral multi-temporal pairs of images. Further to the
visual interpretation, the study illustrates a spectral enhancement method using a PCA.
The results show some clear changes in the urban fabric in southern area of the walled
city of Nicosia, helping to identify several significant ‘hot-spots’. Possible use of such
624 B. Cuca et al.
result could be seen in the development and implementation of the urban masterplans
or planning the construction works that could have high impact on the historic urban
structure e.g. new high speed transport connections, proximity of larger constructions
to sensitive areas such as Venetian walls and so forth. However, it is important recall
that in order to quantify and describe the changes more precisely, further in-situ
measurements and comparison with historic information such as maps and aerial
images needs to be performed. In such a way, it would be possible correlate in an even
more significant manner the observations coming from satellite imagery with changes
in the urban scenario that have occurred over the past three decades.
Acknowledgements. The present communication is under the “ATHENA” project

H2020-TWINN2015 of European Commission. This project has received funding from the
European Union’s Horizon 2020 research and innovation programme under grant agreement
No. 691936.
References
1. Neocleous, N., Papantoniou, A.: The implementation of the DLS portal and the INSPIRE
geoportal of Cyprus. In: Proceedings of Fourth International Conference on Remote Sensing
and Geoinformation of Environment (RSCy2016), Pafos, Cyprus (2016)
2. Janssen, K.: Open government data and right to information: opportunities and obstacles.
J. Community Inf. 8(2) (2012)
3. Cuca, B.: Geospatial future is open: lessons learnt from applications based on open data. In:
Gervasi, O., Murgante, B., Misra, S., Rocha, A.M.A.C., Torre, C., Taniar, D., Apduhan,
Bernady O., Stankova, E., Wang, S. (eds.) ICCSA 2016. LNCS, vol. 9788, pp. 491–502.
4. Directive 2003/98/EC on the re-use of public sector information (2003)
5. Agapiou, A., Alexakis, D.D., Lysandrou, V., Sarris, A., Cuca, B., Themistocleous, K.,
Hadjimitsis, D.G.: Impact of urban sprawl to archaeological research: the case study of
Paphos area in Cyprus. J. Cult. Heritage 16(5), 671–680 (2015)
Towards the Sustainable Development
of Cultural Landscapes Through Two Case
Studies on Different Scale
Eirini Papageorgiou(&)

[email protected]
Abstract. Since the beginning of the twentieth century the definition of Cul-
tural Heritage has gradually expanded from the scale of individual monument to
the scale of cultural landscapes. The broadening of the term has at the same time
increased the complexity of the information originating from different domains
and being on different scales and forms. In this context, the objectives as well as
the challenges involved in the Cultural Heritage sector have become highly
diversified, often leading to fragmented and less successful interventions which
do not conform with the principles of Sustainable Development. Therefore
Cultural Heritage and Sustainable Development should correlate with each
another. Pursuing the achievement of sustainable models of development for
cultural landscapes, this paper investigates how the factor of scale can act as a
linkage between the fields of Cultural Heritage and Sustainable Development.
Keywords: Cultural landscapes River valleys Network of settlements

Scale factor Sustainable development Heritage management Strategic
planning Cultural Heritage
1 Introduction
1.1 Key Considerations

The concept of a historic monument embraces not only the single architectural work
but also the urban or rural setting in which is found the evidence of a particular
civilization, a significant development or a historic event. This applies not only to great
works of art but also to more modest works of the past which have acquired cultural
significance with the passing of time [1]. The above definition indicates that a number
of characteristics are incorporated in the concept of Cultural Heritage (CH) which relate
to time, space, scale, as well as tangible and intangible properties [2, 3]. The concept of
CH nowadays refers not only to a variety of constructions offering physical evidence of
the past, but embraces also places and surroundings [4]. This expansion of the concept
allows not only monuments, groups of buildings and sites [5] to be considered as CH
assets, but also urban centers, natural, mixed and cultural landscapes [6]. The gradual
shift of the concept of CH asset from individual building to group of buildings,
buildings’ surroundings, districts, sites and whole areas indicates a parallel shift from
the scale of the individual building and building complex to the scale of urban, town

DOI: 10.1007/978-3-319-48496-9_50
626 E. Papageorgiou
and regional planning. In this context, the bigger the scale, the bigger the integration of
“environment” in the concept of CH gets. In addition to the concept of “environment”,
the Burra Charter clearly states that cultural significance is embodied in the place itself,
providing emphasis on the physical location of a place, which may have a range of
values [7]. Landscape is considered to be an autonomous entity and a key element for
the achievement of sustainable development, based on a balanced and harmonious
relationship between social needs, economic activity and the environment [8].
Futhermore, the man-made environment correlates with the natural one, within it has
been created and evolved, as well as with the society which has generated this CH asset
on the one hand by providing funds and creativity and on the other hand by depending
its livelihood on the beneficial uses of it. In this context, CH is regarded as one of the
four interdependent elements, namely economic, social, cultural and environmental
[9, 10].
On the other side, environment, society and economy are the three pillars, on the
common grounds of which lies “Sustainable Development”. According to the European
Commission1, sustainable development stands for meeting the needs of present gen-
erations without jeopardizing the ability of future generations to meet their own needs
– in other words, a better quality of life for everyone, now and for generations to come.
It offers a vision of progress that integrates immediate and longer-term objectives,
local and global action, and regards social, economic and environmental issues as
inseparable and interdependent components of human progress. In this context, CH is
an element of Sustainable Development, as an integral part of each one of the envi-
ronmental, social and economic components of the latter; a part that should be pro-
tected and transmitted to the future generations to guarantee their development [4].
Therefore CH acts at the same time both as an asset to be protected and supported
in order for its existence to be safeguarded through time, and as a provider of the tools
and framework which help to shape, delineate and drive the development of tomor-
row’s societies [11]. This twofold role of CH and its bidirectional relation to the
parameters of environment, society, economy and the overall concept of Sustainable
Development are illustrated in the figure below (Fig. 1).
Since the role of CH has been recognized to be an active one and due to its
multifaceted nature, the need for CH management has been created [4, 8]. This man-
agement is also necessary to be assessed under the prism of sustainability, and towards
that direction different methodological approaches have been developed [10, 12–15].
At the same time the role of strategic planning, as the process of designing a desired
future and identifying ways to bring about it [18], has been highlighted [16, 17].
1.2 Case Studies – Selection and Goals

The large amount of information included in the field of CH, its multilevel interrelations as
well as the complexity of the objectives and the challenges it faces, make any
decision-making process difficult, often leading to fragmented and less beneficial
1
http://ec.europa.eu/environment/eussd/ (accessed: 09/2016).
Towards the Sustainable Development of Cultural Landscapes 627
Fig. 1. Cultural Heritage and Sustainable Development (S.D.): a relation of mutual contribution
interventions. The bigger the scale of the CH asset, the bigger gets the problem of managing
it and conform the interventions applied upon it to the principles of sustainability.
Through the consecutive presentation of two real case studies of cultural landscapes
in mountain regions, which differ in scale, this paper aims to investigate the research
question whether it is possible to achieve sustainable development on each one of these
two scales of CH independently or only through the parallel and complementary survey
of both spatial scales, in the context of regarding space and its components as a
“whole”.
The first case study refers to a network of nineteen riparian rural settlements along
the axis of the river Karpenisiotis in Evritania prefecture (central Greece). The second
case study refers to a single traditional settlement situated in the mountain region of
Zagori in Epirus (northwest Greece). Both of the surveyed areas represent the combined
works of nature and man [5], strongly influenced by the geomorphology, which has
played a catalytic role in the definition of their boundaries as well as in the formation of
their character. They are illustrative of the evolution of human society and settlement
over time, under the influence of the physical constraints and/or opportunities pre-
sented by their natural environment and of successive social, economic and cultural
forces, both external and internal [6], thus are characterized as cultural landscapes.
The purpose of this paper is the correlation of two multiparametric fields (CH,
Sustainable Development) though an approach that extends in time, space, scale and
properties, by using a variety of tools originating from different scientific domains. The
goal of the aforementioned approach is to maximize the management capacity of the
complex and diversified data of cultural landscapes, secure the efficiency of the
research outcomes and safeguard the sustainability of the proposed interventions.
2 Methodology
2.1 The Method

Attempting to comprehend the character of the surveyed area the research in both case
studies was conducted in four phases. In these phases the data from the sectors of
628 E. Papageorgiou
environment, society and economy were collected, analyzed, assessed and formulated
into proposals. Each phase included several steps, sectoral and intersectoral ones. In
this context the parameters of time, space, scale and properties were taken into con-
sideration. The bidirectional shift from past to future, from individual building to
region, from microscale to macroscale and from tangible to intangible was a key
element across all phases and all steps of the way, as depicted in Fig. 2.
Fig. 2. Data management in correlation with time, space, scale and properties.
The first phase was the data collection phase, in which all the related documents
and data that describe the spatial and economic evolution as well as the social changes
through time were retrieved through research in archives and various sources. In the
course of this procedure the engagement and the active participation of the local key
players has been proven catalytic for the completeness of all information. Also during
this phase extensive field work has been carried out.
The second phase was concerning the analysis of the collected quantitative and
qualitative data. All the past studies and surveys as well as the implemented actions and
projects were also analyzed, in order to understand the successes but also the failures
and deficiencies of the existing policies. In this phase the combination of various
information management tools which facilitate the depiction of the extensive thematic
information, such as the ones used in the domain of architecture, urban and regional
planning, have been proven to be highly supportive.
Following the phase of analysis, in the next phase the data were assessed also with
the use of tools originating from the business sector, which have facilitated the clas-
sification of the latter into groups of strengths, weaknesses, opportunities and threats.
In the fourth phase, the information retrieved from the previous stages was used to
formulate the proposals.
In any case, the aforementioned data that have been collected, analysed and
assessed, can be considered as an “information hub” about the region (Fig. 2); an
“information hub” which can support any decision-making process in terms of strategic
planning. This “information hub” can also act as a fertile platform, upon which a pool
of knowledge [19] can be created, in the meaning that all the involved parties (i.e.
stakeholders, policy makers, local communities etc.) can complementarily and coher-
ently create their own knowledge and added value for the region.
The above described methodology is presented in the following Sects. 2.1.1 and
2.1.2 in more detail for each one of the selected case studies. The steps taken as well as
the tools used are different, depending on the scale of the surveyed area. In the case of
the network of settlements, the scale extends from the one of regional planning to the
one of urban planning, while the case of the individual rural settlement is presented
through the shift from the scale of urban planning up to the scale of individual building.
2.1.1 Network of Settlements - from the Scale of Regional Planning

to the Scale of Urban Planning
Environment. In Step 1, the boundaries of the surveyed area were defined according
to specific criteria (Fig. 3 in Sect. 3).
In Step 2, data were gathered concerning the natural characteristics of the area such
as climate, fauna and flora, aquatic resources, geomorphology, environmental zones,
etc.
In Step 3, data about the typology of each settlement on the basis of urban planning,
the building capacity and the land uses were gathered.
In Step 4, data regarding the built environment were collected, such as general
architectural typology of the buildings, important constructions (bridges, watermills
etc.) as well as significant locations and paths around the settlements.
In Step 5, the correlations between the natural and the built environment were
surveyed.
In Step 6, data were gathered which concern the infrastructures.
In Step 7, a survey about the past and current administrative organization of the
area and the prefecture has been conducted (Fig. 5 in Sect. 3), as well as the past and
current legislative framework that is related to the urban and regional development.
In Step 8, data about the interconnections of the settlements are gathered, such as
paths and trails that connect them etc.
630 E. Papageorgiou
Society. In Step 1, data about the characteristics of the population were gathered such
as gender, age, schooling, density etc., as well as their change within a time span of a
decade, in order to understand the nature of the inhabitants and their potential con-
tribution to the actions proposed (Fig. 4 in Sect. 3).
In Step 2, data were gathered concerning historical facts, local traditions, festivals
etc.
Economy. In Step 1, data regarding the three main production sectors (primary, sec-
ondary, tertiary) were gathered.
In Step 2, data regarding the labor status of the population were collected, as well as
their change within a time span of a decade.
Intersectoral (Environment – Society – Economy). In Step 1, an extended analysis
of all past studies and surveys, as well as of the implemented projects and actions has
been conducted, for documentation as well as for assessment reasons.
In Step 2, there has been a contact with the representatives of the local communities
as well as with the ones of the local and central authorities, in order to document their
view regarding the problems, the deficiencies and the needs as well as the possible
solutions. Moreover, in order to take into account their future vision for the area.
In Step 3, the Strengths and the Weaknesses of the region were highlighted
alongside with the Threats that it faces and the Opportunities to be considered.
In Step 4, historic data were detected and interrelated with the partial abandonment
of the settlements (Society-Environment) and the negative impact on Economy.
2.1.2 Individual Settlement - from the Scale of Urban Planning

to the Scale of Individual Building
Environment. In Step 1, the already existing boundaries of the surveyed area were
highlighted.
In Step 2, data were gathered regarding the natural environment, namely geomor-
phology, climate, fauna and flora, rainfalls, protected environmental areas etc.
In Step 3, data about the built environment were gathered, concerning the classi-
fication of the settlement within the wider network of settlements to which it belongs.
In Step 4, data about the settlement’s typology on the basis of urban planning, its
building capacity and the land uses were gathered.
In Step 5, built and non-built plots, built and free spaces as well as their indicators
were defined (Fig. 9 in Sect. 4). Also the building uses were identified. Moreover, the
role of landscape in the formation of the plots, of the urban units and of the road system
was surveyed.
In Step 6, the architectural plans of the dwellings and of other special buildings of
the settlement (church, school etc.) were collected and their form, design, construction
materials and bioclimatic attributes were detected.
In Step 7, data regarding the connection of the settlement with the surrounding area
were gathered, such as networks of paths, bridges etc.
Society. In Step 1, data about the characteristics of the population were gathered, such
as gender, age, schooling, density etc., as well as their change within a time span of the
last decade, for the same reason mentioned in Sect. 2.1.1.
Economy. In Step 1, data regarding the labor status of the population were collected,
as well as their change within a time span of the last decade.
Intersectoral (Environment – Society – Economy). In Step 1, historic data were
detected, in particular the ones interrelated with colonization (Society-Economy) and
the gradual increase of the number of dwellings (Environment), thus concerning the
transformation of the built space through time.
2.2 The Tools

The different scale of the case studies imposed the combined use of different tools, in
order to manage and process the large amount of the collected information. These tools
are briefly presented below:
2.2.1 Network of Settlements - from the Scale of Regional Planning

(1a) GIS, as a spatial analysis tool which facilitates the spatial distribution of various
indicators, (1b) Thematic maps, which illustrate various information deriving from the
phase of analysis, (1c) Overall maps for the surveyed area, (2) Databases, (3) Aerial
photographs (scale 1:30.000), (4) Site photographs, (5) Geophysical maps (scale:
1:50.000), (6) Topographic maps (scale 1:5.000), (7) Charts, depicting the analyzed
data, (8) Documentation sheet for each settlement (Fig. 6 in Sect. 3), with visual and
descriptive information regarding the following data: name of the settlement, munic-
ipality, prefecture, altitude, total area (ha), settlement’s typology, population charac-
teristics and their change within the time span of a decade, special historical data with
spatial impact, tangible and intangible cultural data (festivals etc.), special buildings
and constructions (churches, watermills etc.), special locations and routes close to the
settlements’ boundaries, topographic map, aerial photograph and site photograph of the
settlement, (9) Documents, which include data that can only be described in text form,
(10) SWOT analysis.
In particular, SWOT analysis has been proven to be a very useful tool for the
combination of the multiple information of the surveyed area and has provided a good
basis for the formation of the strategic planning of the latter. In the spatial context of the
particular case study, SWOT analysis has also been supplemented by the use of GIS,
which has enabled the spatial distribution of the indicators under analysis and high-
lighted environmental aspects and human interactions with the natural environment,
thanks to the overlaying of thematic layers [15]. This kind of spatial SWOT analysis is
a tool which has also been used in other case studies [20–22] and includes data from
the phase of analysis, regardless if they can be quantified or not.
632 E. Papageorgiou
2.2.2 Individual Settlement - from the Scale of Urban Planning

to the Scale of Individual Building
(1) Aerial photopraphs (scale 1:5.000), (2) Site photographs, (3) Topographic maps
(scale 1:5.000), (4) Thematic maps, which illustrate various information regarding the
built and the free space of the settlement as well as its buildings (scale: 1:1.500), (5)
Charts, depicting the statistic data, (6) Documentation sheet (Fig. 11 in Sect. 4) for each
one of the buildings of the settlement, with visual and descriptive information regarding
the following: name of the settlement, municipality, prefecture, number of the building
and number of the urban unit to which it belongs, past and present building use, dating
of the building, names of owners and their social characteristics, photograph and
topographic map of the building and its plot, urban planning indicators, photographs and
architectural drawings per elevation, short architectural description, buildings’ typo-
logical characteristics and general architectural typology, construction materials, state of
preservation, evaluation of importance and proposed protection grade, (7) Architectural
drawings, (8) 3D sketches, (9) Typology table (Fig. 10 in Sect. 4), which includes the
classification of the settlement’s buildings into the main identified categories (types and
subtypes), alongside their architectural and dimensional variations.
In the following Sections (Sects. 3 and 4), the two case studies are presented along
with a short description and a sample of the work done throughout the four phases of
data management. This presentation acts complementarily to the facts that have already
been described in Sects. 1 and 2, as well as to the ones that will be analyzed in Sect. 5
(Results).
3 Case Study 1: Network of Settlements – Development

Perspectives for the Settlements of Karpenisiotis River
Valley in Evritania Prefecture
The nineteen riparian settlements along the axis of Karpenisiotis river, belonging
nowadays to three different municipalities, constitute a unity of special interest that has
been surveyed in this dissertation of 2004 [23]. In particular, the river is situated at an
altitude zone of 500 m.– 600 m. above sea level while all settlements are situated at an
altitude zone between 700 m. and 1000 m. above sea level (Fig. 3).
In addition to the landscape characteristics, according to which a zone has been
defined, the network of the settlements is established by the dense mesh of man-made
paths that unites them and also from many historical facts, such as that in 1836 A.D.
these settlements used to belong to the same municipality, which means that the area
was considered to be homogeneous. Due to their location, their proximity to the river
and the river valley which supported the existence of the settlements and the habitants’
living, as well as their common characteristics in various other fields of tangible and
intangible CH, these settlements are recognized to be a unity with common problems
and needs and therefore be treated as a network with similar development potentials.
The boundaries of the network are defined in such way, in order to align to geomor-
phological characteristics (natural boundaries) (Fig. 3), thus surpassing the existing
Fig. 3. Identification of the surveyed riparian area (valley of Karpenisiotis river)
ones, which have been defined according only to strict authority and administrative
criteria (artificial boundaries) (Fig. 5).
Target of this project is the sustainable development of the area along the
Karpenisiotis river through the judicious management of its resources and potentials.
Fundamental belief is that the development of this (river) axis can serve as a model for
the contiguous riparian valleys and their settlements, thus disseminating the develop-
ment in the whole region. Worthmentioning is that all collected information was found
scattered in various archives and authorities, while this project managed to gather the
multiple and multifaceted information into one place, acting for the first time as an
information and knowledge hub [24] about the region.
634 E. Papageorgiou
Fig. 4. Population data and their changes within a decade (1991–2001)

Fig. 5. Administrative structures and correlations in different time periods (1836, 1984, 2004)
Fig. 6. Typical settlement’s documentation sheet

636 E. Papageorgiou
Fig. 7. Strategic axes for the development of the wider region (below, right); Strategic
masterplan with axes, targets and actions specifically for the surveyed area and funding
opportunities for their implementation (center); Detail of a proposed target with its supplemen-
tary actions (below)
4 Case Study 2: Individual Settlement - Conservation

and Redevelopment of the Traditional Rural Settlement
of Koukouli in Zagori of Epirus
After the study of a “network of settlements”, the scale shifts from regional to urban.
This case study refers to a single rural settlement with already defined boundaries and
was surveyed in the context of a M.Sc. dissertation in 2006 [25]. The settlement of
Koukouli in Zagori of Epirus (northwest Greece) is a traditional village situated on
Tymfi mountain at an altitude of 880 m. above sea level, close to the Vikos gorge
(Fig. 8) and the valley of Voidomatis river. It is surrounded by outstanding natural
environment, which is protected by national and international laws. Moreover, it
belongs to a wider network of settlements in the territory, called ‘Zagorochoria’, which
according to national laws is recognized to be a network of settlements carrying ancient
traces of regional, urban and spatial organisation that have to be safeguarded.
Fig. 8. The Vikos gorge in the wider Fig. 9. Urban units and indicators about the built
region (Copyright: Geozah (CC BY-SA and non-built space of Koukouli (Figure created by
4.0, https://commons.wikimedia.org/w/ the author; site plan adapted by [26])
index.php?curid=49173760)
Worthmentioning in this case is that the existing administrative boundaries of the

settlement (illustrated in Fig. 9 with a strict outline) do not comply neither with the
formation of the landscape nor with the boundaries that are distinctively formulated by
the outline of the existing urban plots.
638 E. Papageorgiou
Fig. 10. Typology table with schematic representation of the dwellings’ plans alongside their
architectural and dimensional variations, classified in three basic categories (upper part);
Schematic 3D sketches representing the evolution and variations of the dwellings’ volume types
(lower part). (Figures, schematic 2D-3D representations and concept created by the author;
dwellings’ survey plans adapted by [26])
Fig. 11. Typical building’s documentation sheet (Figure created by the author; dwellings’
survey plans adapted by [26])
Fig. 12. Map of the proposed protection grade of the buildings, following the preservation state
assessment in 2005 (Figure created by the author; site plan adapted by [26])
640 E. Papageorgiou
Fig. 13. Proposed town plan (Figure created by the author; site plan adapted by [26])
Fig. 14. Proposed architectural interventions on urban scale

5 Case Studies Results
Depending on the scale of the area surveyed in each case study, the research results
differ and are presented below:
5.1 Network of Settlements - from the Scale of Regional Planning

In Step 1, a strategic masterplan for the development of the region was defined, as
illustrated in Fig. 7 (see Sect. 3). Seven strategic axes were specified, each one of them
was subdivided into targets and each one of the targets was further on subdivided into
distinct actions.
In Step 2, funding opportunities were detected for all proposed actions included in
the masterplan, thus confirming that the latter one is applicable within the current
financing framework.
In Step 3, a series of thematic maps were produced, which illustrated the suggested
spatial interventions, such as transportation network, infrastructures, environmental
interventions, interventions for touristic development etc.
The aforementioned maps were complemented by a book of 220 pages, which
included a full text description of the research procedure in all phases and the steps
followed, as well as of the proposed strategic plan.
5.2 Individual Settlement - from the Scale of Urban Planning to the Scale
of Individual Building
In Step 1, a new detailed Building Regulation was defined, aligned with the actual state
of the particular settlement, its building and land use capacity and development
potentials. This new Building Regulation was formulated specifically for the particular
settlement, in contrast to the existing general Building Regulation which was the same
for all the settlements belonging to the network of ‘Zagorochoria’, without having
taken into account the particularities of each one of them. The proposed Building
Regulation included also guidelines for the restoration of the old traditional buildings
and the permitted interventions upon them according to their protection grade, the
construction of new buildings in order to comply to the character of the settlement,
construction materials, urban equipment etc.
In Step 2, a new town plan was proposed, which included all the building and land
uses, the interventions on urban scale, the redefinition of the settlement’s boundaries
etc. (Fig. 13 in Sect. 4).
In Step 3, various thematic maps were produced in order to illustrate the spatial
interventions, regarding the protection grade imposed on the buildings (Fig. 12 in
Sect. 4), the planning of public spaces etc. (Fig. 14 in Sect. 4).
The data about the settlement throughout the phase of collection, analysis,
assessment and proposal alongside with documentation sheets (Fig. 11 in Sect. 4),
642 E. Papageorgiou
maps and charts have been included and extensively described in a book of 240 pages.
This book supplemented in text form all the aforementioned (data and procedure).
Furthermore, the town plan alongside with the Building Regulation were the
operational guidelines of the strategic planning, to which any intervention conducted
within the settlement, should be aligned.
6 Conclusion
In both aforementioned case studies the exhaustive procedure of data collection,

analysis, assessment and proposal in the context of time, space, scale and
tangible-intangible properties in each one of the fields of environment, society, econ-
omy and culture has provided exceptional input for the strategic planning of each one
of the two regions. Despite the fact that the same procedure was followed on each one
of the two different scales (same four phases, same parameters of time, space etc. taken
into consideration), diversified outcomes and distinct strategic planning proposals have
been developed, which have subsequently led to different kind of interventions on each
scale. Looking at the bigger picture, the interventions proposed on one scale stop at the
point where the interventions of the next scale begin and do not disrupt each other.
Minor overlapping of them on the urban scale can only be regarded as a “buffer zone”
which enables the assimilation of the interventions and the smooth bidirectional shift
from one scale to the other.
Through the case studies presented, it has become clear that the survey of every
settlement individually is, in principle, positive, because all the data that have con-
tributed to the formulation of the settlement and its character are gathered and assessed,
its particularities are identified and therefore taken into consideration along the process
of decision making and planning of the interventions upon it. However, for the
interventions not to be fragmented and for the development of each settlement to be
actually successful and beneficial, its planning should be incorporated into the planning
of the broader spatial network to which it belongs (network of settlements), and
reverse. Both of these scales act complementarily, enhancing each other with data and
safeguarding the efficiency of the decisions taken on each scale, thus ensuring
sustainability.
Under this prism the answer to the research question whether it is possible to
achieve sustainable development on each one of these two scales of cultural landscapes
independently, regarding each case as individual case study, or through the parallel and
complementary survey of both spatial scales, in the context of regarding space and its
components as a “whole”, has also become clear.
Acknowledgments. The research leading to these results has received funding from the People
Programme (Marie Sklodowska-Curie Actions) of the European Union’s Seventh Framework
Programme FP7/2007-2013/under REA grant agreement no. 608013. Special thanks also to my
supervisors Dr. Eleni Andrikopoulou, Professor of Architecture, Dr. Nikos Papamichos, Assis-
tant Professor of Architecture and Dr. Michael Nomikos, Assistant Professor of Architecture at
the Aristotle University of Thessaloniki (AUTH) for the scientific support and guidance
throughout the aforementioned projects.
References
1. ICOMOS: International Charter for the conservation and restoration of monuments and sites
(The Venice Charter) (1964)
2. ICOMOS: The Nara document on authenticity (1994)
3. Vecco, M.: A definition of cultural heritage: from the tangible to the intangible. J. Cult.
Heritage 11, 321–324 (2010)
4. UNESCO/ICCROM/ICOMOS/IUCN: Managing cultural world heritage (2013)
5. UNESCO: Convention concerning the protection of the world cultural and natural heritage.
In: General Conference Seventeenth Session, vol. 1, pp. 135–145 (1972)
6. UNESCO: Operational guidelines for the implementation of the world heritage convention
(2015)
7. Australia ICOMOS: The Burra Charter (1999)
8. Council of Europe: European landscape convention (2000)
9. Vernieres, M., Patin, V., Mengin, C., Geronimi, V., Dalmas, L., Noel, J.-F., Tsang King
Sang, J.: Methods for the economic valuation of urban heritage: a sustainability-based
approach (2012)
10. Dalmas, L., Geronimi, V., Noël, J.F., Tsang King Sang, J.: Economic evaluation of urban
heritage: an inclusive approach under a sustainability perspective. J. Cult. Heritage 16, 681–
687 (2015)
11. ICOMOS: 17th general assembly and scientific symposium “Heritage, driver of develop-
ment”. ICOMOS News, vol. 18 (2011)
12. Ferretti, V., Bottero, M., Mondini, G.: Decision making and cultural heritage: an application
of the multi-attribute value theory for the reuse of historical buildings. J. Cult. Heritage 15,
644–655 (2014)
13. Ferretti, V., Comino, E.: An integrated framework to assess complex cultural and natural
heritage systems with multi-attribute value theory. J. Cult. Heritage 16, 688–697 (2015)
14. Oppio, A., Bottero, M., Ferretti, V., Fratesi, U., Ponzini, D., Pracchi, V.: Giving space to
multicriteria analysis for complex cultural heritage systems: the case of the castles in Valle
D’Aosta region, Italy. J. Cult. Heritage 16, 779–789 (2015)
15. Comino, E., Ferretti, V.: Indicators-based spatial SWOT analysis: supporting the strategic
planning and management of complex territorial systems. Ecol. Indic. 60, 1104–1117 (2016)
16. Cerreta, M., Panaro, S., Cannatella, D.: Multidimensional spatial decision-making process:
local shared values in action. In: Murgante, B., Gervasi, O., Misra, S., Nedjah, N., Rocha,
A.M.A.C., Taniar, D., Apduhan, B.O. (eds.) ICCSA 2012. LNCS, vol. 7334, pp. 54–70.
17. Fusco Girard, L., De Toro, P.: Integrated spatial assessment: a multicriteria approach to
sustainable development of cultural and environmental heritage in San Marco dei Cavoti,
Italy. Cent. Eur. J. Oper. Res. 15, 281–299 (2007)
18. Steiner, G.: Strategic planning: what every manager must know, a step-by-step guide (1979)
19. Žarnić, R., Rajčić, V., Moropoulou, A.: Identity card of cultural heritage: how to collect and
organize data. In: Ioannides, M., Fritsch, D., Leissner, J., Davies, R., Remondino, F., Caffo,
R. (eds.) EuroMed 2012. LNCS, vol. 7616, pp. 340–348. Springer, Heidelberg (2012).
doi:10.1007/978-3-642-34234-9_34
20. Baycheva-Merger, T., Wolfslehner, B.: Evaluating the implementation of the Pan-European
criteria and indicators for sustainable forest management – a SWOT analysis. Ecol. Indic. 60,
1192–1199 (2016)
644 E. Papageorgiou
21. Geneletti, D., Bagli, S., Napolitano, P., Pistocchi, A.: Spatial decision support for strategic
environmental assessment of land use plans. a case study in southern Italy. Environ. Impact
Assess. Rev. 27, 408–423 (2007)
22. Sanò, M., Fierro, G.: Integration of the SWOT analysis as a coastal management tool with a
geographical information system: two approaches to the problem and first results
Dipartimento per lo studio del Territorio e delle sue Risorse (Dip. Te. Ris.) Università di
Genova. Univ. Genova (2003)
23. Papageorgiou, E.: Development perspectives for the settlements of Karpenisiotis river valley
in Evritania prefecture, Dissertation, AUTH (2004). (Original in Greek)
24. Ioannides, M., Davies, R., Chatzigrigoriou, P., Papageorgiou, E., Leventis, G., Niko-
lakopoulou, V., Athanasiou, V.: 3D Digital Libraries and their contribution in the
documentation of the past. In: Mixed Reality and Gamification for Cultural Heritage.
25. Papageorgiou, E.: Conservation and reformation of the traditional rural settlement of
Koukouli in Zagori of Epirus. M.Sc. dissertation, AUTH (2006). (Original in Greek)
26. Christidis, V.: The architecture of the village of Koukouli in the area of Zagori, Epirus, vol.
1, 2, Rizareio Foundation, Athens (2004) (Original in Greek)
The Fortification System of the City of Urbino:
The Case Study of Valbona Gate
from 3D Surveys to GIS Applications
for Dynamic Maps
Sara Bertozzi, Laura Baratin(&), and Elvio Moretti
Conservation and Restoration School – DiSPeA,

University of Urbino Carlo Bo, Urbino, Italy
{sara.bertozzi,laura.baratin,elvio.moretti}@uniurb.it
Abstract. The city walls of the city of Urbino, originally Metaurense Urvinum,
today represent the result of an evolution that has antique origins, from proto-
historic settlements to the first Roman ruins, up to its current conformation,
expression of the important Renaissance period. We present a study of docu-
mentation, survey and analysis of the various parts of the walls, analysed in their
entirety and in individual elements, among these we focus on the main access
gate to the historic centre, Valbona Gate. The integration of avant-garde tech-
nologies during all the phases, from the survey to the graphic rendering, up to
advanced management with GIS instruments allows us to have a picture of the
architectural reality both of the present and the past, fundamental for scheduled
maintenance and for planning future conservation interventions. The geometric
and morphological survey brought the process to completion with the problems
relating to deterioration, the understanding of the situation of the pathologies
and the descriptive dimension of the architectural elements in a complete bi- and
tri-dimensional GIS system that allows us to take advantage of a complete series
of processing and statistical assessment capabilities. The project is part of a
wider program of research on the walled city of Urbino, wherein the analysis of
the city walls integrates with the diachronic analysis of its expansion and the
geomorphological context in which it is found as well as a 3D City Model and
the analysis and management of the built environment, current and
archaeological.
Keywords: 3DGIS Walled city Urbino
1 Introduction
The management, evaluation and maintenance capabilities of the architectural heritage

are currently based on the immense capacity of reading and interpretation of knowledge
concerning the use of innovative computing instrumentation. In particular, a funda-
mental concept that is exalted by the integrated use of organic surveying techniques
with those for the management of data is that of documentation. It serves as a tool to
make information accessible to those who cannot investigate the property itself and it
becomes an integral part of the conservation process that should be based on thorough
DOI: 10.1007/978-3-319-48496-9_51
646 S. Bertozzi et al.
research, physical and analytical surveys [1]. The study of a walled city provides a
thorough assessment of its historical evolution and the relationship that the architectural
composition and its development may have had with the very life of the city and its
inhabitants. It is therefore necessary to define and organise all the information con-
cerning both the internal living apparatus, which constitutes the historical core of the
city, and the walls themselves that delimit it, defensive structures that over the centuries
became a historical testimony and physical delimitation of the cultural and vital centre
of the city. The walls had functioned as a protection for the inside inhabitants from
outsiders, identified the city’s boundary, and in some cases defined powerful families’
or rulers’ territory [2]. The entire documentation process involves a painstaking phase
of identification and recording of the information that is useful in defining an appro-
priate planning strategy for conservation and intervention. The complex nature of the
historic city requires a detailed documentation and the mapping of its cultural and
natural characteristics, a thorough knowledge of its urban morphology, historical
evolution, physical and functional characteristics and socio-economic processes.
Several works have been carried out on the analysis of the walled cities and on city
walls, for a deeper understanding of the city and to preserve elements of such great
importance [3–5]. For some years now in the city of Urbino, UNESCO World Heritage
Site, there has been a series of research programmes that have allowed us to carry out
various levels of investigations [6–9], from the geomorphological evolution of the
territory on which the town is located to a diachronic analysis of the metamorphosis of
the city walls from the preurban phase to the current state, up to a detailed definition of
the urban apparatus through the 3D City Model and an integrated analysis of plans and
elevations of individual architectural buildings of interest [10, 11], all managed in the
GIS environment for a possibility of advanced management.
We now define, in the proposed work, a thorough analysis of the city walls, to
proceed to an operation of documentation and assessment of the possible deterioration
and impairment due to events which have taken place in the course of time, evaluating
the various levels of risk and the applicable interventions. The analysis of the built
environment in elevation can also be carried out through a stratigraphic reading of the
archaeological type, in fact, the sequence with which various parts of the architectural
object were built or restored can be documented with methods similar to those applied
to deposits that have been buried. Indispensable for rendering this process operational
is the high reliability and resolution of the survey, because, even if the artefact is still
the primary source of information, there is the absolute necessity to record every
consideration that derives from direct observation in a scientifically correct way.
Naturally, the three-dimensional model produced by a series of laser scans remains the
best way to implement this virtuous process. Its management through GIS technology
allows us to implement the organisation and the interrogation of the data supplied by
the survey by means of a relational database. To implement the project on the city walls
of Urbino, a series of laser scanner surveys have been carried out, starting from the
main Gates, allowing us to read the individual stones or bricks, determine the various
types or characteristics, establish the conservation status and record all previous
restoration interventions to which it had been subject over time. The result is a dynamic
mapping of the “GIS based” prospectuses made available to the management entities,
in order to provide scheduled maintenance that limits the maintenance or reconstruction
The Fortification System of the City of Urbino 647
interventions over time. In order to implement an effective scheduled conservation

action, it is not enough to have a thorough physical knowledge of the object one seeks
to preserve, but it is necessary to know and analyse historical events connected to it and
to accurately reconstruct all the previous interventions of restoration that the object has
undergone in the course of time.
2 Urbino Walled City
The birth of the Duchy of Urbino dates back to 1443, in virtue of the nomination of
Oddantonio II da Montefeltro to Duke of Urbino by Pope Eugene IV. He reigned for
less than a year. In 1444, he was in fact assassinated following a popular uprising,
which brought his older half-brother Federico to power, who will prove himself to be
one of the greatest princes on the Italian chessboard of the era, as famous for his role as
commander in battle as he was renowned as a cultured patron of the arts. Under his
guidance, from 1444 to 1482, the duchy soon became one of the focal points of the
Italian Renaissance. During this brief period the city underwent a radical transforma-
tion, induced especially by the decision of Federico to build his residence on the
Poggio, already seat of the first settlement in the Roman epoch. The Palace would
become over the centuries a true manifesto of Renaissance architecture and has been
defined by the critic Kenneth Clark as: “the prototype of the Renaissance palaces and
the most beautiful in the world”.
The urban structure that the city assumes after the interventions carried out during
this period and which has come down to us still intact, even now represents the summit
of Renaissance architecture, harmoniously adapted to its physical environment and to
its past, making it an entirely exceptional place, which since 1998 has been registered
in the list of Unesco World Heritage sites. The main actors of urban transformations
were mainly the architects Luciano Laurana and Francesco di Giorgio Martini who
would leave their mark, still visible today in the architectural structure of the city.
In 1508 the duchy passed to the della Rovere family, who transferred the seat of the
duchy to Pesaro and then passed to the State of the Church in 1631. In 1621, on the
occasion of the marriage of the last Duke of Urbino, Federico Ubaldo della Rovere who
married Claudia de’ Medici, Porta Valbona was erected. A new golden age for the city
of Urbino arrived at the beginning of the eighteenth century, following the election to
the papacy of Clement XI, part of the Albani family. This figure, with enlightened
patronage, promoted the construction of civil and religious buildings, contributing to
giving a more modern structure to the city, without however changing the development
of the walls in a significant way. In the course of the 1860s only a small portion of the
wall was demolished in order to build a Toll Barrier, named after the then Princess
Margaret of Savoy, erected in part also on the Bastion of Belisario. This change,
however modest, still resulted in the rebuilding of the section of the walls between the
Bastion of San Polo and the small Bastion of San Girolamo, demolished during these
works. Built at the same time was the new road that, passing along the walls, connects
the Barrier Margherita to the current Piazza della Repubblica square that led to the
closing of the San Polo Gate. In 1906, again with the aim of further improving access to
the city, the Del Monte Gate was demolished and buried or most of the bastion was
knocked down to create a lookout on which the monument to Raphael was erected.
Taking into account these changes, the historic city walls of Urbino, as shown on the
the chorographic plan of 1841 of Fig. 1, include seven city gates, thirteen bastions and
the Albornoz Fortress.
Lastly, in the sphere of the reorganisation and restructuring plan of the historic
centre launched by the city of Urbino in 1964, which included the restoration of the
patrol walkways on the walls, the reopening of the Francesco di Giorgio ramp and the
recovery of the old stables of the Duke called the “La Data”, also included was the
transformation of the Mercatale into underground parking for cars below the level of
the plaza. The works began in 1969 and were concluded in 1972 by the architect
Giancarlo De Carlo. With the parking lot and the closure of the railway line that
connected Urbino to Fano in 1987 the Mercatale saw a vast increase also of the number
of buses arriving, and the square continues to carry out its function as the main access
to the city of Urbino, even if today the bus station has been moved to the complex of
Santa Lucia Gate.
3 Valbona Gate
The most important of the changes that have been imposed following the construction
of Palazzo Ducale was not the demolition of many old buildings located on the Poggio,
where the ancient Roman city once had risen, but the urgent need to acquire new spaces
for the expansion of the entire urban nucleus.
Palazzo Ducale, which exemplifies this theme, stands on the Roman forum (altitude
450 m) but the western side acquires space with an almost vertical development,
conquering the entire escarpment and individuating two fundamental levels, that of the
foundation of the Torricini (small thin towers) (altitude 433 m) and that of Mercatale,
delineated by the Renaissance walls on one side and by the great work that is the
“Risciolo” by Francesco di Giorgio Martini on the other. The huge substructure wall is
composed of seven masonry vaults that support the Mercatale Plaza that, as local
tradition dictates, was created over the rubble and soil that was moved there so as to
build the foundation of Palazzo Ducale itself and that, upon its completion at ground
level, reaches an altitude of 410 m. The base of the substructure wall is located at the
minimum altitude of 384 m and then, in its maximum elevation, covers a vertical drop
of no less than 26 m. The section of Fig. 2 shows the altitude profile starting from the
Albornoz Fortress on the left, to Palazzo Ducale, with the positioning of the square of
Piazzale Mercatale, Valbona Gate, the substructures of the Risciolo and the Spiral
Ramp.
In the vast Piazzale del Mercatale (Market Square), which owes its name to the fairs
and markets that were held there in the past, rises Valbona Gate, the architectural
construction connected to the city walls, designed and built by the Urbino architect,
Sigismondo Albani. It is the only Gate of Urbino with a monumental face directed
outwards, opening toward the main access route to the city. It is an architectural
product that is perfectly up-to-date with the examples of the late Mannerist and
Baroque period, very probably derived from the models of the Treatise by Sebastiano
Serlio, with the face having smooth ashlars to act as a background to the pairs of lateral
Fig. 1. The chorographic plan of the city of Urbino with all the gates, bastions and the Albornoz
Fortress indicated. These artefacts have constituted the defensive apparatus of the city and still
today the majority of them are conserved. Valbona Gate, even today the main point of access to
the city; Santa Lucia Gate, situated on the south-eastern slope of Colle del Monte, north of the
city walls; Lavagine Gate, rebuilt in the seventeenth century, following the demolition of the
homonymous bastion adjacent to the city gate, is located at the bottom of the south-eastern slope
of Colle del Monte; Del Monte Gate, no longer existing, was located on the hilltop of Colle del
Monte, at the summit of the current Via Raffaello, north-west of the ancient city, was demolished
in the early years of the twentieth century in order to improve the viability of the area and access
to the historic center; San Bartolo Gate, located on the eastern slope of Colle del Poggio, east of
the city walls; Santa Maria Gate or Sant’Agostino, located on the western slope of Colle del
Poggio, close to the bastion of Sant’Agostino; San Polo Gate, situated on the southern slope of
Colle del Poggio but, from the 1860s no longer accessible to public transit, after the construction
of the nearby Toll Barrier; the Bastion of Santa Lucia, it is the northernmost tip of the boundary
wall, not accessible to the public, because it hosts a private garden; Bastion del Monte,
semi-destroyed during the early years of the twentieth century, only the base has survived
because it is underground, was adjacent to the homonymous city gate; the Bastion of San
Bartolo, located close to the homonymous city gate, collapsed in February 1928 and rebuilt in
the early 1930s; the Bastion of Santa Chiara, situated on the south-eastern slope of Colle del
Poggio; the Bastion of Belisario, in the south-east of the city walls, was restored in the 1860s, it
conserves part of the buildings of the old Barrier Margherita; the Bastion of San Polo,
representing the extreme southern tip of the walls, was restored in the nineteenth century; the
Bastion of Sant’Agostino, situated on the western side of the walls and of the Colle del Poggio;
the Bastion of Santa Caterina, located on the western slope of Colle del Poggio, rebuilt in 1963,
following a partial collapse due to erosion of the underlying ground; the Bastion della
Santissima Trinità, situated on Colle del Monte, to the north of the city walls between the Del
Monte and Santa Lucia Gates; the Bastion of Lavagine, demolished in the seventeenth century,
was adjacent to the homonymous city gate; the Bastion of San Girolamo, demolished in the
1860s, was a small bastion of semicircular shape on the south-eastern slope of Colle del Poggio,
between the bastions of Belisario and Santa Chiara; the Bastion of the Spiral Ramp, its
principal feature is a large spiral staircase that connects the ducal stables known as “La Data” that
are located adjacent to the section of the walls between this bastion and that of Santa Caterina, at
Palazzo Ducale, (the Ducal Palace). It was designed by Francesco di Giorgio Martini, in the
second half of the fifteenth century, the upper part of this bastion was occupied by some
warehouses, partly demolished in the nineteenth century to build the City Theatre; the Bastion
dei Frati, a small bastion of semicircular shape, on the south-western side of the Colle del
Monte, under the Albornoz Fortress and above Valbona Gate; the Albornoz Fortress, rising in
the south-western corner of the city walls, on Colle del Monte, erected in the second half of the
fourteenth century by Cardinal Anglico Grimoard, modified in several successive stages and
finally restructured in its current form toward the first half of the sixteenth century.
semi-columns supporting the inscription: VRBINVM ROMANORVM ANTI-

QVISSIMVM MUNICIPIVM VMBRIA OLIM VETVSTISSIMA CIVITAS MODO
INTER PICENI MAIORES, LONGE TAME HISCE TEMPORIBVS SVB SEREN-
ISSIMIS DVCIBVS SVIS CLARIOR, SED ILLARIOR NVQ; FEDERICO ET
CLAUDIA PRINCIPIBVS FAVSTVM ET SAECVNDVM PRECATVR CON-
IVNGIVM. The monumental face, deliberately spectacular, was made in 1621, as
indicated by the inscription, on the occasion of the wedding of Prince Federico Ubaldo
della Rovere with Princess Claudia de’ Medici, to welcome the wedding procession
and the bride and groom coming from Florence. Placed along the route were triumphal
arches enriched with figurations, most of them were precarious works. This Gate is the
exception, built with stone and bricks despite the fact that it was known that the
wedding procession and the bride and groom would stay in Urbino just one night. The
Duke established his residence in Pesaro where in February 1622, Vittoria was born,
destined to be the last descendant of Della Rovere family. In fact, Federico Ubaldo
would die the next year in Urbino, having just reached the age of 18, on 28 June 1623.
Still today and not only because it is linked to these historical facts, but also for its
important function as the main access to the city, Valbona Gate is the most important of
the seven gates that are part of the city walls of Urbino.
Valbona Gate has always created many problems linked to deterioration, especially
of the stone material, for which various interventions of restoration were carried out,
indispensable for reaching the current state of conservation. The first was carried out by
Mastro Silvio Tommasini in 1755 and it focused mostly on the ornamental part and on
the stabilisation of the two eagles, work by the architect Giovan Francesco Buonamici.
Included in the context of the same restoration, reinforcements were made to the wall
alongside the Gate. In 1825 it was Giuseppe Corsini who undertook some repairs and
restoration on the external monumental part. It was then the turn of the Technical
Officer of the Municipality of Urbino who, on 27 September 1873 approved a project
for the removal of a public fountain which was close to the door. The reason for the
intervention was essentially of an aesthetic nature because the fountain was used as a
drinking trough for animals. However, a subsequent appraisal dated 17 October 1873
highlighted the severe conditions of deterioration involving the entire gate and he
ensured that work of a structural type of restoration would be carried out with extreme
urgency. Despite the work carried out, in 1784 there was the need once again for a
major intervention of consolidation and maintenance. Also in the 1950s, it was nec-
essary to intervene with other structural work, a result of which was the walkway above
the gate being created, by removing some houses the two ends of via Mazzini were
Fig. 2. Section circa W-E of the city of Urbino which shows the difference in height on which
the main buildings of the city are located: from right to left the Albornoz Fortress, Piazzale
Mercatale with the substructures of Risciolo and Valbona Gate, the Spiral Ramp, Palazzo Ducale
and the complex of Santa Chiara.
connected, as well as other minor changes including the insertion of an epigraph at the
summit of the pediment that testified the restoration work carried out in those years; the
text reads: ANNO JVBILAEI MCML AERE PVBBLICO PRINISTINI OPERIS AD
EXEMPLAR RESTITUTUM. The last restoration interventions in order of time date
back to 1985–1986 and were carried out by the firm Carli and again in 1995 by the
same company on indications of Croci.
4 Survey and GIS Projet
The analysis and management of data relative to the walls of Urbino for different levels
of users takes place through various steps and the integration of a series of method-
ologies and technologies that find in GIS the optimal support for complex and diverse
needs. The initial input data are obviously the base cartographies, in particular the
1:1.000 maps supplied by the Municipal administration, showing all of the elements
that compose the historical centre of Urbino and the areas adjacent to the fortified
system, both as urban fabric (buildings, roads, walls etc.) and as territorial character-
istics (contour lines outside the urban centre, topographic points, boundaries of green
areas, etc.). Integrated with them are the CTR, (Regional Technical Map), having a
1:10.000 scale, thematic geological and geomorphological maps, acquired by the
CARG project, again at the 1:10.000 scale in Raster format, which are the basis for a
reconstruction of the orography of the terrain in current terms and in those of historical
evolution, in addition to maps of various historical eras and the iconographies for the
transformation of the urban area both in the historical centre and along the city walls.
The project concerning the walls of Urbino aims at providing an effective instru-
ment in the framework of scheduled maintenance that can manage not only the
restoration interventions, but also serves as continuous monitoring of the current sit-
uation and a means for future evaluations that also take into account the future val-
orisation of this historical segment of the city.
As indicated, the project is part of a wider programme of research on the man-
agement of the heritage of the city’s historical centre, part of the UNESCO heritage,
both in its current connotations and in those of the past.
The walls are then analysed in their entirety for evaluations of a dimensional and
logistic character by analysing the arrangement of the fortified systems and the entrance
gates to the city, as well as those of a functional nature, analysing some elements of the
fortified structure such as the Bastion of San Paolo, Lavagine Gate, Valbona Gate and
the Albornoz Fortress. The first phase was of course that of acquisition of detailed data
through surveying, photogrammetry and laser scanning for the creation of 3D models
and the restitution of conventional metric drawings: plans, sections and facades. The
initial process is therefore based on a series of surveying operations, aimed at con-
trolling the dimensional consistency, the characterisation of the various descriptive
typologies (history, techniques, materials, form and nature of deterioration, etc.) and to
the creation of the basis on which to operate the subsequent analysis and study of the
themes, to then move to a stage of display and fruition of the data, within which it is
possible to manage a reconstruction of reality in a highly flexible way, that is, one that
is useful for diverse types of end users.
Direct and photographic surveys have led to aerial photography renderings that
shall integrate the topographical survey through total station, carried out through
topographical points identified by architectural targets and the point clouds deriving
from the laser scanner. The entire step of alignment and cleaning of the clouds has led
to the creation of the grid and the 3D model, allowing the insertion of horizontal and
vertical planes that define plans, prospectuses and sections of the final architectural and
metric surveying. The graphic rendering, all the iconographic documentation and
information on the material characteristics and on the various surveys executed in the
course of time on the gate and on the evolution of its conformation, were integrated into
a GIS system (ArcGIS in its latest version 10.3.1) which allows the analysis and
management in two- and three-dimensional environments. The 3D surveys were also
modelled to define the various phases of evolution of the architectural shape of the gate,
starting from the second half of the fifteenth century to its current definition (Fig. 3).
The structuring of the GIS project starts in the 2D environment, then in ArcMap
where a series of methodological steps were defined in order to bring together all the
data available in map, section and prospectus into a single project, connecting them to
one another through complete relational databases, to finally bring about the
three-dimensional environment only as data management and query. The idea is to start
from the georeferenced maps, in this case in Monte_Mario_Italy_2 (WKID: 3004;
Authority: EPS; Projection: Transverse_Mercator; False_Easting: 2520000.0;
False_Northing: 0.0; Central_Meridian: 15.0; Scale_Factor: 0.9996; Lati-
tude_Of_Origin: 0.0; Linear Unit: Meter (1.0), for a general and territorial framework
of the elements being studied. Particular attention is paid in this phase to the organi-
sation of the walls in the map, organised as polygonal features but subdivided into
various segments that identify the areas targeted for intervention to which the reliefs
refer. This subdivision already allows dimensional analysis of the structural elements of
the fortifications and the first statistical evaluations to be performed, bringing about
their connection with the rest of the in-depth information through the organisation of a
relational database. Inserted in the attributes table is an identifier field of the particular
section of the wall, which is used as a key field in a Join operation with prospectuses
and sections of detail, always managed within the same project, but spatially placed in
a completely different position. The GIS is in fact used as a management programme of
Fig. 3. 3D reconstruction of the various building phases that have affected Valbona Gate. The
first gate was built in conjunction with the construction of the walls of the Federico period at
2,67 m from the current level of the road surface, whose foundations have been brought to light
during some excavations. On the occasion of the extension works of Palazzo Ducale, the walls
and the gate were rebuilt, however defining a very simple form with a vault bricks and without
ornaments. On the occasion of the wedding of Federico Ubaldo della Rovere with the Princess of
Tuscany Claudia de’ Medici the gate was adorned with a giant order of twin columns in Tuscan
style, two niches with statues, a frieze and a swan-neck pediment. Over the next three centuries a
series of restoration works were carried out that had not radically modified the monument, with
the exception of two eagles in massif limestone in the first half of the eighteenth century, in place
of the shrines placed above the niches, whose statues had been stolen between the late
seventeenth and early eighteenth centuries. In the course of the centuries, various dwellings were
built against the rear side of the gate, taking advantage of the spaces where the sentinels stood
guard, during the period of the reign of the Montefeltro family. During the last phase in 1950, an
epigraph was placed at the centre of the swan-neck pediment certifying some restoration of the
artistic type carried out on the Gate. In those years there was also the project which involved the
demolition of the homes built onto the gate for the realisation of the walkway that extended on
the eastern part of the walls until the Spiral Ramp of Francesco di Giorgio Martini.
geographical maps, but its development in relation to spatial elements that cannot
undergo classic georeferencing is still of great interest. In this case, the use of the
workspace, therefore the XY plane, is utilised as a simple work plan to manage met-
rically correct spatial data such as prospectuses and sections which can be applied to all
the typical analyses of the Spatial Analyst tools. This XY plane becomes a hypothetical
XZ plane, as if the elevation could be handled in its vertical plane, so as to allow the
transformation of every element in the 3D feature and then subject each one into a
subsequent georeferencing for management in 3D, in the ArcScene environment, in
their correct geodetic position.
However, the entire phase of organisation and data processing takes place in
ArcMap, so that coexisting in a single project are the georeferenced cartography and
the data arising from the reliefs, prospectuses and sections, brought together in a
separate area of the workspace. For the sake of convenience, a linear feature is used that
starts from the 0.0 coordinates by creating the X axis definition line (considering the
workspace as the XZ plane) on which all polygonal features are placed that describe the
vertical movement of the walls. Prospectuses are kept in different layers, for the ele-
vated weight of the thousands of records that describe, element by element, everything
down to the single stone, but each of them has a field that identifies the correspondence
to the element in a map in the georeferenced cartography. In this way, it goes from the
spatial identification of the area of interest, to the detailed relief. Also included are all
the mappable information relative to historical documents and everything that has
covered the different phases of the relief, the detail vertices for the topographic surveys,
the orthophotos, the photoplanes, etc. All information is stored in the Geodatabase
environment, divided into two parts, one that is georeferenced and the other that is not
(Fig. 4).
Fig. 4. The image shows the methodological approach used with GIS to manage the various
components of the walls of Urbino, taking Valbona Gate as an example. At the top, the
georeferenced walls are divided into segments that identify the elements of interest, listed in the
table of attributes with the field NameID, that serves as a key field for the relational database with
which to move directly to the prospectus of interest, in this case that of Valbona Gate. Each
element of the prospectus is associated with a series of information in a table that can provide a
number of different views, with the creation of dynamic cartography, which also allows the
processing of graphs and statistical analyses. The management and the final querying of the
whole project is carried out, after appropriate placing and georeferencing, in the 3D environment,
where maps, models and prospectuses coexist.
Of essential importance was an accurate analysis phase of the types of construction

materials and of the current deterioration that, in some portions of the walls, such as
that from the Albornoz Fortress to the Bastione dei Frati, has brought about outlining a
mapping of the surface deterioration, of the infesting weeds and of the biological
colonisation, leading to identifying the effects that these have had on the masonry, the
major causes that may have led to certain critical issues and the types of intervention
needed to restore optimum conditions. The percentage of coverage of the different
biological forms in the various sections analysed and a thorough assessment of the
characteristics and impact of each of these on the masonry types present has allowed us
to identify a danger index, indicative of the areas on which to intervene with greater
urgency. Sampled and analysed, by means of the mass spectrometer, were also a series
of in-depth samples. Summary data sheets were created to describe the spectra present
and the main descriptive information, transforming them into images to insert in the
attributes table. A special layer, a feature class, indicates the mapping of the samples,
indicating in the database the type of biological species referenced and returning a field
in Raster format that allows the insertion of images that can be directly queried by the
geometric element in the workspace.
5 Conclusions
The project brings about an accurate definition of the walls of Urbino, in particular
Valbona Gate, through a well-defined process that allows us to carry out a detailed
study to acquire knowledge about the historical, artistic and architectural characteristics
of the artefact and its evolution over the centuries, arriving at a complete plan for
currently surveying and managing these characteristics. The integration of advanced
instrumentation allows us to combine the geometric and morphological data with
descriptive qualitative information that leads to the use of final management in two- and
three-dimensional GIS environments. Thanks to the capacity of organising all data into
a single Geodatabase system, the information relating to the conservation status and to
deterioration, restoration interventions carried out, comparisons with other similar
types of gates, the constructive evolution, the mapping of the pathologies, previous and
scheduled interventions, all layers become searchable and stackable, so that they may
be subjected to queries and statistical analysis, the basis for a thorough scheduled
maintenance.
Acknowledgements. We acknowledge students and professors of the First Level Master

“Instruments and Methodologies for Cultural Heritage Conservation and Valorization” of the
University of Urbino Carlo Bo for your cooperation and Municipality of Urbino for all available
materials for research purposes.
References
1. Husseini, B., Bali, Z.: Documentation, using GIS techniques in conservation of a World
Heritage Site, a case study of “The Old City of Jerusalem”. In: The International Archives of
the Photogrammetry, Remote Sensing and Spatial Information Sciences, 25th Interna-
tional CIPA Symposium 2015, Taipei, Taiwan, 31 August – 04 September 2015, vol.
XL-5/W7 (2015)
2. Bennison, A.K., Gascoigne, A.L.: Cities in the Pre-Modern Islamic World: The Urban
Impact of Religion. State and Society. Taylor & Francis, Abingdon (2007)
3. Shah, K.: Documentation and cultural heritage inventories, case of the historic city of
Ahmadabad. In: ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial
Information Sciences, 25th International CIPA Symposium 2015, Taipei, Taiwan, 31 August
– 04 September 2015, vol. II-5/W3 (2015)
4. Mohareb, N., Kronenburg, R.: Arab Walled Cities: investigating peripheral patterns in
historic Cairo, Damascus, Alexandria, and Tripoli. In: Proceedings of Eighth International
Space Syntax Symposium, Santiago, PUC (2012)
5. Guley, K., Abbasoglu, S.: Proposal for the revitalisation of the Walled City Famagusta. In:
2005 WSEAS International Conference on Environment, Ecosystems and Development,
Venice, Italy, 2–4 November, pp. 90–95 (2005)
6. Bertozzi, S., Baratin, L., Moretti, E.: L’evoluzione delle trasformazioni di un centro storico
tramite GIS: la città di Urbino. In: Proceedings of ESRI Italia 2016, Rome, Italy, 20–21
Aprile 2016
7. Baratin, L., Bertozzi, S., Moretti, E.: Le trasformazioni della città di Urbino durante il
periodo dei Montefeltro: tecniche innovative per lo studio delle trasformazioni urbane. In:
Processi di analisi per strategie di valorizzazione dei paesaggi urbani. I luoghi storici tra
conservazione e innovazione, Proceedings. 29 Gennaio 2016, Rome, Italy. A cura di
Gerardo Maria Cennamo. Ermes. Servizi editoriali integrati S.r.l., Ariccia (RM), pp. 105–
115 (2016). ISBN 978–88–6975–094-6
8. Baratin, L., Bertozzi, S., Moretti, E.: GIS intelligence for a cutting-edge management of 3D
Cities. In: Proceedings of Digital Heritage – International Congress 2015, Granada, Spain,
28 Settembre – 2 Ottobre 2015 (2015)
9. Baratin, L., Bertozzi, S., Moretti, E.: The Geomorphological transformations of the City of
Urbino: the design of the city analysed with GIS tools. In: SCIRES it, SCIentific RESearch
and Information Technology, vol. 5, no. 1, pp. 41–60 (2015). e‐ISSN 2239‐4303, doi 10.
2423/i22394303v5n1pp41 © CASPUR-CIBER Publishing, http://caspur-ciberpublishing.it
10. Bertozzi, S., Moretti, E.: Gestione integrata dei rilievi del Portico di San Francesco ad
Urbino. In: Baratin, L, Acierno, M, Muratore, O. Strumenti e Metodi per la Conservazione e
la Valorizzazione dei Beni Culturali. Ancona, IL GABBIANO S.r.l., pp. 114–121 (2015).
ISBN: 978‐88‐905347‐7‐5
11. Bertozzi, S., Baratin, L., Moretti, E.: Cultural Heritage: restituzione reale di prospetti e piante
in ArcScene. In: Proceedings of 15 Conferenza Italiana Utenti ESRI, Rome, Italy 09–10
aprile 2014, Supplemento a GEOMEDIA, vol. 2–2014, ISSN: 1128-8132
12. Mazzini, F.: I mattoni e le pietre di Urbino, pp. 62–64, Argalia editore, Urbino (1982)
13. Luni, M.: Archeologia nelle Marche - Dalla preistoria all’età tardoantica, pp. 195–196,
Nardini editore, Firenze (2003)
Project Paper: Innovative Methods on
Risk Assessment, Monitoring and
Protection of Cultural Heritage
Disaster-Risk Indicators and Their Evaluation,
Communication for Traditional Settlement
Alex Ya-Ning Yen and Chin-Fang Cheng(&)
China University of Technology, 56 Sec. 3 ShingLong Rd., Taipei 116, Taiwan

{alexyen,aabbyy}@cute.edu.tw
Abstract. Disaster-risk management has become an important issue in the

conservation of cultural heritage since the beginning of 21st century. However
the implementation procedure of the disaster-risk management was mainly
focused on single monument and short of the research for the settlements which
is still a big challenge for us. In 2005, to comply with the international trend,
Taiwan amended the Cultural Heritage Conservation Act, in which settlements
as a new category of monuments were declared – “Settlement: a group of
building s, street houses, settlements which have architectural style, landscape,
historical, artistic or scientific value.” This research develops disaster-risk
indicators for the traditional settlement and takes Quion-lin settlement, a World
Heritage potential site as an example, analyzing the procedure to set up a net-
work of conservation value and the disaster-risk indicators for traditional set-
tlement. GIS will play as an important tool for the integration and
communication within the stakeholders.
Keywords: Disaster-risk indicators Traditional settlement Cultural

heritage GIS
1 Introduction
In 1989, the General Assembly of the United Nations proposed a natural disaster
reduction plan (IDNDR, 1990–1999) for next decade. Yokohama Strategy Plan In 1994
released an action program on the prevention, preparedness as well as mitigation for the
prevention of natural disasters. In 2005, United Nations Department of International
Strategy for Disaster Reduction (UNISDR) announced the Hoygo Framework for
Action 2005–2015, HFA, for disaster mitigation extending to the level of national and
community resilience. The overall goal was trying to establish a global strategy for
disaster mitigation.
In the beginning 20 years, the objects of world cultural heritage sites were mainly in
single buildings. However, under the influence of the Washington Charter, 1987,
Cultural landscape Charter, 1992, Xian Declaration, 2005 and Quebec Charter, 2008,
the concept of conservation has extended to the site and place of heritages. The con-
servation of Historic Urban Landscape had been noted as an important challenge in the
“Operational Guidelines for the Implementation of the World Heritage Convention,
OG.” A management approach methodology has been discussed in recent years and the
disaster-risk management has been identified as one of the important parts of this

DOI: 10.1007/978-3-319-48496-9_52
660 A.Y.-N. Yen and C.-F. Cheng
approach. However, the tools on the risk assessment and monitoring for the imple-
mentation of traditional settlements are important research issues.
In 2005, to comply with the international trend, Taiwan amended the Cultural
Heritage Conservation Act, and declared settlements as a new category of monuments.
“Settlement: a group of buildings, street houses, settlements which have architectural
style, landscape, historical, artistic or scientific value”.
This research takes Quion-lin settlement, a World Heritage potential site as an
example, analyzing the procedure to set up a network of conservation value and the risk
indicators for traditional settlement. GIS will play as an important tool for the inte-
gration and communication within the stakeholders.
2 Literature Review
2.1 Disaster-Risk Indicator

Based on the concept for disaster-risk management of cultural heritage, some inte-
gration framework had been established by important international organizations which
is a top down idea and helps to construct a more comprehensive network with the
relevant disaster prevention system. Under the above effort1, some useful tools were
developed to help the evaluation. 5 important impacts for the World Heritage, (pressure
from the developing, environment, disaster-risk, tourist/recreation and the number of
resident) were identified in the “Format for the nomination of properties for inscription
on the World Heritage List, 2015”. In 2011, the ICCROM had released a toolkit for the
risk assessment and 14 risk indicators were noticed in this document. Which are
buildings and Development, transportation Infrastructure, utilities or Service Infras-
tructure, pollution, biological resource use/modification, physical resource extraction,
local conditions affecting physical fabric, social/cultural uses of heritage, other human
activities, climate change and severe weather events, sudden ecological or geological
events, invasive/alien species or hyper-abundant species, management and institutional
factors, and other factor(s).
2.2 Public Participation

The World heritage Committee released the Budapest Declaration in 2002. A 4C
strategy (Credibility, Conservation, Capacity building and Communication) was
announced to help the conservation of the WH. In 2007, the 31st World Heritage
Committee was held in Christ Church City, New Zealand. In this meeting the fifth C
(Community) was added and became a 5Cs strategy. This issue was put in the article 26
of OG (Fig. 1).
1
Ya-Ning Yen, Chin-fang Cheng, 《Research on the GIS as a communication platform in the risk
management of traditional settlement》, 2014.
Disaster-Risk Indicators and Their Evaluation 661
Fig. 1. 5Cs strategy
The ICOMOS Charter for the Interpretation and Presentation of Cultural Heritage
Sites, 2008, made clear definition and provided efficiency for the general public to
understand the CH.
• Interpretation refers to the full range of potential activities intended to heighten
public awareness and enhance understanding of cultural heritage site. These can
include print and electronic publications, public lectures, on-site and directly
related off-site installations, educational programmer, community activities, and
ongoing research, training, and evaluation of the interpretation process itself.
• Presentation more specifically denotes the carefully planned communication of
interpretive content through the arrangement of interpretive information, physical
access….… and multimedia applications and websites.
2.3 The Experiences of Taiwan

A value based conservation and management approach were key concepts of the
Conservation Act, 2005; public participation was identified as an important part of
implementation. Based on this Act, a self-management plan for the site manager was
developed. Through a 4 years’ research, 18 potential risk indicators of the monument
were identified and put in the management plan. However, a boarded scope research on
the settlement and environment is still needed.
3 Execute Process
Firstly, from the literature review, this research will establish the risk-disaster indicators
for the historic settlement. Secondly, through AHP method (Analytic FDM (Fuzzy
Delphi Method)) we can set up a working table for evaluating those indicators. To the
end, under the concept of 5Cs strategy, a GIS based cultural mapping will be developed
as a tool for the communication within the stakeholders.
3.1 Risk-Disaster Indicators

There are 2 different types of risk-disaster indicators needs to prepare for certain
purposes. According to the existing research, indicators and the assessment weighting
can be established for single monument. Take 5 topics of the “Format for the nomi-
nation of properties for inscription on the World Heritage List (2015)” as a reference
and make comparison with the 14 disaster-risk indicators by the ICCROM (2009)2, 18
indicators in Taiwan (Table 1)3; we can find the requirements of the WH monitoring
are far beyond the practical needs especially in the management affairs. Based on the
result of Table 1, a similar tool was developed for the reviewing of entire settlement by
AHP (Table 2). Basically, the table includes 3 main parts together with indicators for
the risk assessment: disaster prevention, rescue and recovering.
3.2 Public Participation and GIS

This research takes 5 steps for the implementation (Fig. 2).
Fig. 2. Public communication, participation and GIS
• Build up a basic documentation of the settlement and transfer all the information
into the GIS as a basic platform on site surveying.
2
These 14 indicators were taken in the ICCROM workshop on the monitoring of WH in Suzhou,
China, 2009.
3
Ya-Ning Yen, Chin-fang Cheng, Hung-Ming Cheng, 《Disaster Risk Management and Measure-
ment Indicators for Cultural Heritage in Taiwan》, 2015.
Table 1. Disaster-risk indicators

Format for the nomination of ICCROM (2009) Disaster-Risk indicators in
properties for inscription on the World Taiwan (2016)
Heritage List (2015)
1. Development Pressures 1. Buildings and 10. Disaster history
Development 11. Improper placement of
2. Transportation items
Infrastructure
3. Utilities or Service
Infrastructure
2. Environmental pressures 4. Pollution 16. Pests or ants Road
5. Biological resource 17. Increased moisture, wall
use/modification seepage serious
6. Physical resource
extraction
7. Local conditions
affecting physical
fabric
3. Natural disasters and risk 10. Climate change 18. Other factor(s)
preparedness and severe weather
events
11. Sudden ecological
or geological events
12. Invasive/alien
species or
hyper-abundant
species
4. Responsible visitation at World 2. Transportation 18. Other factor(s)
Heritage sites Infrastructure
3. Utilities or Service
Infrastructure
9. Other human
activities
5. Number of inhabitants within the 8. Social/cultural uses 1. The use of fire
property and the buffer zone of heritage 2. Improper placement of items
9. Other human 3. Undated environmental
activities clean-up
4. Excessive electrical
facilities, lines of old
(including extension cords)
(continued)
Table 1. (continued)
Format for the nomination of ICCROM (2009) Disaster-Risk indicators in
properties for inscription on the World Taiwan (2016)
Heritage List (2015)
13. Management and 1. The use of fire
institutional factors 2. Improper placement of items
3. Undated environmental
clean-up
4. Excessive electrical
facilities, lines of old
(including extension cords)
5. Management organization
and lack of staff
6. Failure in regularly
attending relevant courses or
lack of certificates
7. Cooperation with
community
8. Lack of fire-fighting
equipment
9. Lack of monitoring alarm
12. Insufficient entrances
(including those are often
closed)
13. Unused to be
repaired/repair/repaired
14. Gutters are not smooth
15. Doors and windows
damaged
14. Other factor(s) 18. Other factor(s)
• Invite the stakeholders joining a cultural mapping exercise mainly on the identifi-
cation of the cultural value (tangible, intangible and their locations), disaster history
and potential risk. All the results will integrate into GIS.
• Survey and evaluate all the possible risk-disaster indicators of the monuments,
settlement and mark in the GIS.
• Discuss with the stakeholders on the results of step 2, 3 which process can help
them to get community awareness on the risk preparedness and the necessity of its
management.
• Establish a cooperative based on the information from the GIS of the GOV and the
community. At this moment, GIS will become a tool to help the management,
education, communication and monitoring.
Table 2. Risk assessment in settlement

A. Disaster prevention
Nature, (1) Climate Change, (2) Typhoon, (3) Flooding, (4) Earthquake,
Environment (5) Fire, (6) Biological, Microorganism
Manmade (1) Military, (2) Fire, (3) Burglar, (4) Traffic accident
disasters
Development (1) Illegal development, (2) Traffic
pressures
Usage behavior (1) Tourist, (2) Improper use, (3) Festival
Mitigation (1) Exercises, (2) Education
B. Disaster rescue
Evacuation (1) General evacuation, (2) Tourists evacuation
Roadside Assistance (1) Main road, Secondary roads, (2) Evacuation Site
Natural resources (1) Water, (2) Well
Equipment (1) Fire hydrant, (2) Alarm device, (3) Monitoring, (4) Water Supply
C. Recovering
Physical (1) Resettlement sites, (2) Materials, (3) Craftsman, (4) Administrative
Support
Psychological (1) Tutor, (2) Resettlement
4 Quion-Lin Settlement Example
4.1 Background
Kinmen is one of the 18 potential sites of World Heritage sites in Taiwan with “various
threats of war in 1600 years’ history in-between continuing to pursue the values of
peace” as its major value. Quion-lin settlement is one of the main strongholds of the
Kinmen inscription. Tsai clan based in Quion-lin settlement has more than 1,000 years
of history and many ancestors in the family became government officials or made
glorious military achievement. Currently the settlement has about 1,800 residents, 8
national monuments, 1 county cultural settlement, 2 historic buildings, 2 intangible
cultural heritages (Wind Lion God, Quion-Lin Ancestor worship), 332 traditional
settlements, accounting for 43.92 % of all, is known as a living traditional culture
settlements. Quion-Lin settlement was registered as cultural settlement in December
2011.
4.2 Procession
This research takes 5 steps to implement the survey of disaster-Risk indicators and their
assessment, communication with the stakeholders (Fig. 3). After the advanced survey
and analysis, a disaster- risk map (Fig. 4) was developed together with several
important issues.
Fig. 3. Procession of disaster-risk indicators
Risk indicators
Aggregate
indicators
Risk analysis
Fig. 4. Disaster-risk map
1. People’s understanding of the disaster mostly restricted to individual homes.

Important factors relevant whole settlement, such as spreading risk, relief roads, as
well as emergency water was lack of pervasive awareness.
2. GIS is an efficient tool for risk communication and as a platform for the advanced
researches’ implementation such as the simulation by FDS, PATH FINDER etc.
4.3 Discussion
• Through a 3 years’ processing in Quion-lin (2013–15), a clear risk management

plan has been established which can help the advanced researches. However the
following implementation framework such as legal, training, monitoring and inte-
grative exercise is still needed (Fig. 5).
Fig. 5. Popular participation in safety management system
• In this case, the residents of Quion-lin can understand the importance of risk
management. On the other hand, how to balance the contradiction within the
conservation and developing pressure is still a big challenge.
5 Conclusion
In the era of 21 century, the challenge for the conservation is not only to proceed with
the monument or its culture content, but also focuses on the additional value resulted
from public participation, education and management during the process of conser-
vation. There are many researches on the enforcement of awareness for public par-
ticipation in conservation. Various topics have been discussed in those papers, such as
method, tool, medium, value, assessment, etc. for risk management. Moreover, how
these efforts can be applied on site?
This research finds,
• By referring the existing document and experiences, we can establish the
risk-disaster indicators and the relevant information of traditional settlement as a
foundation for building the risk management plan. From onsite surveying, assess-
ment (monuments, entire environment) and the message from the cultural mapping
with the stakeholders can make a great help to integrate various ideas. GIS is an
efficient tool in this process.
• Disaster prevention and management should not only relay on some physical
monitoring of cultural heritage (such as electronic devises’ monitors), but also
strengthen cooperation between the public and private sectors. Integrating the
information from public participation, the GIS system can play as a communication
platform for risk management as to linking the private community with other public
departments to make added value.
Acknowledgements. Special thanks to the project “Research on the risk communication and
resilience of historic district” (MOST 103-2625-M-163-001), sponsored by the Ministry of
Science and Technology.
References
1. Klir, G.J., Folger, T.A.: Fuzzy Sets, Uncertainly and Information. Prentice Hall, Upper Saddle
River (1988)
2. ICCROM, Wijesuriya, G.: Approaches to condition assessment (2009)
3. China University of Technology: Kinmen County Settlement King Lam preservation and
redevelopment plan, Kinmen County Department of Cultural Affairs (2014)
4. China University of Technology: Monuments built and settlements district calendar
professional services centre (II), Bureau of Cultural Heritage (2016)
5. UNESCO: Operational Guidelines for the Implementation of the World Heritage Convention
(2015)
6. Yen, Y.-N., Cheng, C.: Research on the GIS as a communication platform in the risk
management of traditional settlement. In: Euromed 2014, Cyprus (2014)
7. Yang, W.-B., Cheng, H.-M., Yen, Y.-N.: An application of G.I.S on integrative management
for cultural heritage- an example for digital management on Taiwan Kinmen cultural heritage.
In: Euromed 2014, Cyprus (2014)
8. Yen, Y.-N., Cheng, C., Cheng, H.-M.: Disaster risk management and measurement indicators
for cultural heritage in Taiwan. In: 25th International CIPA Symposium (2015)
9. http://www3.rits-dmuch.jp/anshinanzen/. Accessed May 2016
Pervasive Wireless Sensor Networks
for the Monitoring of Large Monumental
Structures: The Case of the Ancient
City Walls of Siena
Alessandro Pozzebon(B) , Tommaso Addabbo, Ada Fort, Marco Mugnaini,

Enza Panzardi, and Valerio Vignoli
Department of Information Engineering and Mathematical Sciences,

University of Siena, Via Roma 56, 53100 Siena, Italy
[email protected],
{addabbo,ada,mugnaini,panzardi,vignoli}@ii.unisi.it
http://www.iism.unisi.it
Abstract. In this paper, a solution for the pervasive monitoring of large

monumental structures based on the use of Wireless Sensor Networks is
presented. In particular, the paper focuses on the case of the Ancient
City Walls of the city of Siena, Italy, that still surround the whole his-
toric centre and require a real time monitoring of the cracks present in
several points.
Two different network topologies are presented for the deployment
of a pervasive monitoring infrastructure, and a novel sensing platform
based on the use of Hall effect based sensors is presented. The architec-
ture of the whole sensor node is described, together with the laboratory
test phase that proves the effectiveness of the proposed solution. The
proposed solution is expected to be deployed in a 1.8 Km section of the
city walls in the next months.
Keywords: Wireless Sensor Network · Monumental structures · Hall

Effect sensor · Mesh network · Distributed network
1 Introduction
The monitoring of the static of vast monumental buildings poses several problems
in terms of system design, pervasive coverage and data reliability. With the term
“Monumental buildings” we indicate all these structures that are characterized
by large dimensions and a high historical, artistic and cultural value, like castles,
cathedrals or palaces. These buildings can be in some cases centuries old, and
present structural cracks that can evolve leading to permanent damages or even
collapses.
The monitoring of these structures and the realization of alert systems is a
complex activity due to many factors. The biggest one is the dimension: large

DOI: 10.1007/978-3-319-48496-9 53
670 A. Pozzebon et al.
buildings may present several critical points, thus requiring a large quantity of
sensing devices. This means that the realization of a pervasive infrastructure can
be expensive and depends on the number of sensors to be deployed. Moreover,
data acquisition can be complex with a wired sensor network: reaching all the
sensors with a cable would mean to cover the surface of the building with a wide
network of wires.
The second factor that affects the deployment of a sensing infrastructure in
a large monumental building is its high historical and artistic value: this implies
that the instrumentation to be positioned on the structure must have a low visual
impact. Small sensors should be employed, possibly choosing devices that could
be embedded inside the walls. Similarly, wirings should be avoided or minimized.
Finally, the last factor that must be taken in account concerns the effective-
ness of the monitoring infrastructure as an alert system to warn against possible
risk situations. This means that the monitoring infrastructure should be able to
provide data in real time and then remotely: commonly used visual systems like
crack meters require the intervention of an operator to manually detect possible
changes in a crack and may not be sufficient to detect possible collapses.
Wireless Sensor Networks (WSN) [1] can represent the ideal solution to face
all the problems listed before. The term WSN is widely used to indicate all
the monitoring infrastructures composed by a variable number, from few unities
to even some thousands, of sensing devices, called Sensor Nodes, provided with
wireless connectivity, and then able to transmit, locally and remotely, all the col-
lected data sets. Sensor nodes can be deployed in large quantities, thus covering
even large structures, they can be designed to reduce as much as possible (even
down to some tens of euros) their cost, and their wireless connectivity prevents
from the installation of wires. Sensor nodes are usually small, and their shape can
be designed to be adapted to the final installation environment, thus reducing
as much as possible their visual impact. Finally, Sensor Nodes can be provided
with global connectivity, allowing the real time data transfer to a remote data
collection centre in charge of providing alerts in case of critical situations.
The use of WSNs for Structural Health Monitoring has been widely studied
[2,3]. The most part of the described systems focus on the monitoring of civil
infrastructures [4]: for example in [5] the use of WSN in bridge health monitoring
is described, while in [6] the use of intelligent WSNs fior the evaluation of concrete
buildings is presented. Some solutions focusing on the use of WSNs for the
monitoring of historical buildings can also be found [7,8]: in particular, in [9] a
Wireless Sensor Network for the monitoring of the medieval Torre Aquila in the
city of Trento, Italy is described, while [10] presents a WSN to be deployed in the
Basilica S. Maria di Collemaggio in the city of L’Aquila, Italy, badly damaged
by the 2009 earthquake.
The solution proposed in this paper focuses on the monitoring of an historic
structure of huge dimensions, the City Walls of the city of Siena, Italy, and pro-
poses the use of novel, high-precision, low-cost sensors, that can be employed in
large quantities, thus allowing the deployment of a truly pervasive monitoring
infrastructure. The paper is structured as follows: in Sect. 2 the case study, i.e.
Pervasive Wireless Sensor Networks 671
the City Walls of Siena, are briefly presented, together wit their peculiar struc-
tural problems. Section 3 describes the proposed Network architectures, while
the structure of the sensing device is presented in Sect. 4. Section 5 presents the
result of the first tests carried out in the last months while in Sect. 6 some final
conclusions are presented together with the work to be carried out in the next
future.
2 The Case Study

As already told, the proposed monitoring infrastructure has been shaped to be
deployed on the ancient City Walls of the city of Siena, Italy (Fig. 1). This huge
architecture still surrounds the whole historic centre of the city and is one of the
most imposing witnesses of the importance of the city during the Middle Ages.
It was built along several centuries, even if the walls that can be seen today were
mostly built in the thirteen and fourteen century. The path of the walls runs
for about 7 km, enclosing an area of about 180 ha: it is interrupted only by 8
monumental gates. The structure reaches in some points a height of 15 m, and
a thickness of about 2 m.
While the whole structure is currently in a good state of preservation, it
presents several hundreds of cracks that require continuous monitoring to avoid
possible collapses which may also occur as a result of critical meteorological
phenomena. In particular, the Municipality of Siena is currently focusing its
attention on a 1.8 Km section of the walls in the south-east corner of the city,
moving from the Porta Romana monumental gate to the Porta Ovile monumen-
tal gate. This is the section that will be interested by the first deployment of
the monitoring infrastructure and where a test campaign has been started in
August, 2016. In this section, two sites have been identified for the first tests,
the Porta Romana test site (Fig. 2(a)) and the Porta Ovile test site (Fig. 2(b)):
in these sites cracks moving along the three dimensions can be found, requiring
the design of three different sensor layouts. Moreover, this section is surrounded
by green areas, without connection to the electricity line: this means that an
energy harvesting solution has to be set up to allow the continuous operation of
the system.
3 Network Architecture
In order to develop a monitoring infrastructure able to cover the whole section of
the City Walls under study, two different network topologies have been proposed,
all based on the use of IEEE 802.15.4 XBee radio modules, providing a maximum
transmission range of around 100 m in line of sight. While the thickness of the
Walls could reduce this range, all the sensors are expected to be positioned on
the same side of the walls: this means that they are ideally always on line of
sight.
The first network topology is a centralized architecture based on a mesh
topology, while the second one is a distributed topology composed by a set of
Fig. 1. The historic centre of Siena with the City Walls.
(a) (b)
Fig. 2. The two test sites: Porta Romana (a) and Porta Ovile (b).
autonomous sub-networks based on a star topology. These two different topolo-

gies are due to the main features of the structure to be monitored: the City
Walls extend in length and the deployment section is around 1.8 Km long, and
then a centralized star topology has to be discarded due to the impossibility
to reach the farthest nodes. Similarly, a tree topology was discarded due to the
high dependence on failures to intermediate nodes.
The first topology (Fig. 3(a)) foresees the realization of a single Gateway
node provided with both local (IEEE 802.15.4) and global (GPRS) connection.
This node is in charge of receiving all the data transmitted by the other nodes
and transmit them to the remote data collection centre. The nodes are arranged
according to a mesh topology based on ZigBee protocol, that allows multi-hop:
this solution is especially fault tolerant and it allows the operation of the system
in the event of failure of a single node. On the other side, the system is strictly
dependent on the operation of the Gateway node: if this node fails, the whole
infrastructure can no longer operate.
The second topology (Fig. 3(b)) is based on the deployment of more than
one Gateway, in charge of collecting the data from all the nodes in line-of-
sight around it. Such a solution has higher costs due to the need of more than
one GSM transmission module, but is more resistant to failures: in particular,
the autonomy of the single star networks allows each sub-system to operate
autonomously and not depend on the functioning of the other sub-systems.
4 Sensor Node Structure
The wall crack monitoring is performed by a low-cost dedicated displacement

sensor developed by the authors that is based on a commercial Hall sensor. The
device is based on a permanent magnet placed in front of an integrated Hall
sensor and insures the required measurement range of 4 mm and an accuracy of
10 µm.
Figures 4 and 5(a) show the crack monitoring node and the sensor structure
respectively.
In Fig. 5(b) the sensing principle is illustrated: the permanent magnet
mounted on one side of the crack generates the magnetic flux density field which,
far from its surface on its axis, can be approximated by the following equation:
2μ0 μ
Bx = , (1)
4πx3
where x is the distance, μ is the magnetic field moment and μ0 is the magnetic
permeability of vacuum. The Hall sensor, mounted on the opposite side of the
crack, trasduces this field into a Hall voltage, namely V outsens , which is related
to the magnetic flux density field B as follows:
V outsens = KH IBx , (2)

where KH is the Hall constant, I is the current in the sensor, and Bx is the
component of the magnetic flux density perpendicular to current flowing in the
sensor.
Since the overall displacement provides a voltage output non-linearly related
to the distance x between the magnet and the sensor surface, the working point
(a)
(b)
Fig. 3. The two network topologies: Centralized mesh topology (a) and Distributed
star topology (b).
Fig. 4. Schematic diagram of the overall measurement system.

Fig. 5. Structure of the displacement sensor for crack monitoring (a). The linear circuit
to adapt the sensor output voltage range to the input full scale of the ADC in Fig. 4(b).
was carefully selected in order to grant a large sensitivity and a sufficient mea-
surement range. The Hall voltage is amplified with a linear conditioning system
(Fig. 5(b)), providing the voltage
Vout = (VH − Vref )A0 , (3)
where A0 is the overall gain and Vref is an offset voltage. The ADC resolution
sets the basic constraint to the overall design, whereas the voltage full-scale sets
a limit to the distance range (xM IN , xM AX ).
5 Test and Validation

The proposed proof of concept system shown in Fig. 4, has been designed using
the Hall-effect sensor A1302 by Allegro Microsystems and a neodymium mag-
netic disk (NdFeB) with dimensions 8 mm (height) and 10 mm (diameter), and
a magnetization grade N45, suitable to operate up to 80 ◦ C. The output volt-
age range of the Hall sensor has been adapted to the input full scale 0−5 V
range of a 10−bit ADC converter. It is important to mention that the variation
of the environmental temperature becomes an issue in a long-term monitoring
of the crack width. Indeed, temperature variations affects both the response of
the electronic devices and mechanical structure of the sensor causing possible
alterations in the measurements. For this reason, the sensing system has been
equipped with an integrated-circuit for temperature sensing, namely the LMT84
by Texas Instrument. This sensor provides an analog output voltage proportional
to the temperature in the range from −50 ◦ C to 150 ◦ C.
The Hall sensor and the magnet are coupled by a 3D printed ABS holder
(see Fig. 6(a)), which allows for fixing the magnet and the Hall sensor to the
two sides of the crack, maintaining the alignment and housing also the front end
electronics and the temperature sensor.
The developed displacement sensor was characterized by varying the distance

x in Fig. 5(a) within the range 0–30 mm.
In Fig. 7 some experimental results show the overall dependence of the dis-
placement sensor output on the temperature. The data are obtained by keeping
a fixed distance between the Hall sensor and the magnet and letting the temper-
ature vary. It is possible to observe that this dependence is fairly linear, therefore
the information given by the temperature sensor can be easily used to compen-
sate for the measurement drift of the displacement sensor. The proof of concept
monitoring system was tested placing two sensors on field. In Fig. 8 some mea-
surement results are shown. Data are related to a measurement time of about
two days (measurements are transmitted with time intervals of 1 min) and come
from a sensor mounted on the city wall in the month of August. In the upper
plot both raw distance data (blue line) and data compensated for temperature
(red line) are reported.
Fig. 6. Printed ABS holder for the magnet, the Hall and temperature sensors (a).
Printed circuit boards hosting the sensors and the conditioning electronics (b).
Fig. 7. Dependence of the sensor output on the temperature. The distance between
the magnet and the Hall sensor was kept at a fixed distance and the temperature was
varied.
Fig. 8. Distance and temperature variations measured by the device under test (blue
line). The compensation of the effect of the temperature variations on the displacement
measurements is reported in red line. (Color figure online)
6 Conclusion
In this paper the realization of Wireless Sensor Node based on the use of Hall
Effect sensor for the high precision monitoring of crack displacement in monu-
mental structures has been presented. The proposed solution has been tested in
laboratory providing the required accuracy. At the same time, a network archi-
tecture has been studied to allow the deployment of the proposed node in large
quantities in order to set up an alert system covering all the structure to be
monitored. While the architecture has been thought for a specific deployment
site, it can be easily adopted for any kind of situation where the monitoring of
large monumental structures is required.
The laboratory tests proved the effectiveness of the solution. A test campaign
has been started in August and sensors have been placed in two test sites. The on
field test is preliminary to the deployment of the whole monitoring infrastructure
that is expected to be completed in 2017.
References
1. Yick, J., Mukherjee, B., Ghosal, D.: Wireless sensor network survey. Comput. Netw.
52(12), 2292–2330 (2008)
2. Federici, F., Alesii, R., Colarieti, A., Graziosi, F., Faccio, M.: Design and validation
of a wireless sensor node for long term structural health monitoring. In: 2013 IEEE
SENSORS, pp. 1–4 (2013)
3. Torfs, T., Sterken, T., Brebels, S., Santana, J., van den Hoven, R., Spiering, V.,
Zonta, D.: Low power wireless sensor network for building monitoring. IEEE Sen-
sors J. 13(3), 909–915 (2013)
4. Yun, C.B., Min, J.: Smart sensing, monitoring, and damage detection for civil
infrastructures. KSCE J. Civil Eng. 15(1), 1–14 (2011)
5. Zhou, G.D., Yi, T.H.: Recent developments on wireless sensor networks technology
for bridge health monitoring. In: Mathematical Problems in Engineering (2013)
6. Amditis, A., Stratakos, Y., Bairaktaris, D., Bimpas, M., Camarinopolos, S.,
Frondistou-Yannas, S., Torfs, T.: An overview of MEMSCON project: an intelli-
gent wireless sensor network for after-earthquake evaluation of concrete buildings.
In: Proceedings of 14th European Conference Earthquake Engineering (2010)
7. Capella, J.V., Perles, A., Bonastre, A., Serrano, J.J.: Historical building monitor-
ing using an energy-efficient scalable wireless sensor network architecture. Sensors
11(11), 10074–10093 (2011)
8. Rodriguez-Sanchez, M.C., Borromeo, S., Hernndez-Tamames, J.A.: Wireless sensor
networks for conservation and monitoring cultural assets. IEEE Sens. J. 11(6),
1382–1389 (2011)
9. Zonta, D., Wu, H., Pozzi, M., Zanon, P., Ceriotti, M., Mottola, L., Corr, M.:
Wireless sensor networks for permanent health monitoring of historic buildings.
Smart Struct. Syst. 6(5–6), 595–618 (2010)
10. Potenza, F., Federici, F., Lepidi, M., Gattulli, V., Graziosi, F., Colarieti, A.:
Long-term structural monitoring of the damaged Basilica S. Maria di Collemaggio
through a low-cost wireless sensor network. J. Civil Struct. Health Monit. 5(5),
655–676 (2015)
Project Paper: Digital Applications for
Materials’ Preservation and
Conservation in Cultural Heritage
The SACRE Project: A Diagnosis Tool of Built Heritage
Sarah Janvier-Badosa, Kévin Beck, Xavier Brunetaud, and Muzahim Al-Mukhtar ✉

( )
PRISME Laboratory UPRES EA 4229, University of Orleans,

8 rue Leonard de Vinci, 45072 Orleans, France
{sarah.janvier,kevin.beck,xavier.brunetaud,
muzahim.al-mukhtar}@univ-orleans.fr
Abstract. The SACRE project is a research project which aims to create a tool
to help the professionals in charge of preservation and restoration of cultural
heritage buildings. The objective of this project was to develop a working meth‐
odology and create the technological tools necessary to implement this method‐
ology. This project describes the steps the development of the digital health record
of a building. This project focused on the study and understanding of the mech‐
anisms of degradation of limestone, the main material for building construction.
The Castle of Chambord was chosen to be the subject of this study.
Keywords: Castle of Chambord · Limestone · Degradation · Preservation ·

Digital health record
1 Introduction
In the field of heritage preservation, the practice of restoration of historical monuments

requires to make a diagnosis of the state of conservation before any restore operation.
The SACRE project (Degradation monitoring, characterization and restoration of
limestone monuments) is a research project funded by the Region Centre of France,
between 2008 and 2012. The objective of this project is to provide a tool diagnosis for
preventive conservation and the establishment of a timetable of restorations to preserve
a significant monument: the Castle of Chambord. This project aims at the creation of a
monument health record.
The concept of health records of a monument is an analogy with the medical vocabu‐
lary, which aims to study aspects of the monument health. The health record is both the
support and the result of the diagnosis of the building, as defined by Fitzner [1] and
Smith and Prikryl [2]. The challenge is to synthesize our knowledge of past, present,
and what we can predict of its future.
Creating this health record meets the need of professionals in the conservation and
restoration of historical monuments: know the history of the monument and understand
its current state, in order to predict its evolution. A heterogeneous corpus of historical
and scientific data was collected. The data collected during the study is needed to estab‐
lish a detailed diagnosis of the state of the building, the location of the damage, weather,
historical and architectural archives… The characterization of tuffeau degradation and
understanding of these degradation mechanisms is also a component of this project, as

DOI: 10.1007/978-3-319-48496-9_54
682 S. Janvier-Badosa et al.
a basis for numerical simulation of weathering processes and so in predicting the evolu‐
tion of degradation.
After presenting the administrative and scientific organization of the SACRE project,
this paper will attempt to present the process of constitution of the health record of a
building, by illustrating the methodology with a current type of degradation in the Loire
Valley on tuffeau stone: spalling. A continuation of this part will be the presentation of
the technical tools and specifications for the establishment of a health record that can be
described as “digital”. Finally, this paper will open on different applications and contin‐
uation of this tool and this research project.
2 Project Presentation
2.1 The Castle of Chambord and the Studied Facades
The Castle of Chambord was chosen to be the subject of this project. François 1er begins
its construction in 1519. After several stops and starts of construction, the castle was
partially unfinished at the death of the king in 1547.
His successors will provide various additions, modifications and repairs over the
centuries. The interest of this building is that it is built in tuffeau, the limestone most
commonly used at this period for constructions of the Loire Valley. Posterior restorations
report various types of limestone used instead of tuffeau. In contrast to many buildings
located in urban areas, the Chambord castle is located in rural areas, in the center of a
park of 5500 ha. The study of degraded limestone phenomena is made simpler by freeing
an additional degradation parameter.
The SACRE project focuses on three identified parts of the Domaine national de
Chambord: the South Façade, the East Tower and the Stables of Marshal of Saxe
(Fig. 1). These facades were selected for different characteristics, making them repre‐
sentative of different states of preservation and restoration of the castle. The interest of
the South Façade is that it is the castle entrance facade. It has often been reorganized
and restored over the centuries, and presents a variety of dates and types of stones.
Conversely, for the East Tower, only five of its spans are accessible to visitors, others
have little has been restored and much more damage than the South Facade. Finally, the
Fig. 1. Location of the studied façades in the SACRE project

The SACRE Project: A Diagnosis Tool of Built Heritage 683
Stables of Marshal of Saxe, located near the entrance of the castle, are later, the start of
construction dating from the late seventeenth century. They are now mostly in ruins.
2.2 Objectives and Organization of the Project

The SACRE project was carried by a CNRS laboratory, the CRMD (now ICMN) in
collaboration with the PRISME (laboratoire Pluridisciplinaire de Recherche en Ingéni‐
erie des Systèmes, Mécanique, Energétique), the Domaine national de Chambord and
the LRMH (Laboratoire de Recherche des Monuments Historiques).
This fundamental and applied research program is divided into 5 parts:
1. The CAD modelling of monument aims at constructing a graphic base used to gather
all data acquired during the project.
2. The realization of the health record will reference all the information necessary to
establish a detailed diagnosis of the state of alteration of the monument: mapping of
degradations, weather conditions, architectural and historical archives.
3. The simulation and prediction of degradations, which is the most fundamental step
of this research program, is to simulate the process of degradations both in experi‐
mental aging tests in laboratory and in numerical modelling in order to understand
the evolution and to estimate the kinetics of degradations.
4. The creation of a tool for decision support is the application of simulation, and aims
to estimate the rate of degradation. Added to that a costing of restoration, this soft‐
ware tool will provide a rational schedule of restoration work.
5. The valuation of the project to the public will be achieved by giving an access to a
simplified version of the software, presented at an exhibition at Chambord castle.
3 The Health Record: From the State of the Art to Diagnosis
3.1 Referencing of Historical Data
The first stage of this work was to inventory the historical archives of the castle to find
the different stages of construction and restoration. The data attempting to be accurate
at the scale of the stone. Mappings were performed, which allow to report the extent of
restoration done on these façades. The analysis of this information allows for example
to assess the proportion of visible stones replaced for each of the restoration periods. On
the East tower, nearly 20 % of the stones have been replaced (against 50 % for the South
Façade).
From in situ observations and a summary description of the types of damage, degra‐
dation mappings have also been drawn to scale of the stone (Fig. 2). The spalling was
identified as specific of tuffeau stone, and most destructive. By correlating the mappings
produced, with environmental conditions, it was possible to identify the environmental
factors statistically associated with these degradations. These factors are varied and can
be climatic, anthropogenic and/or related to the nature of the stone and its position in
the building [3]. Textual and graphic data from historical research have completed the
previous analysis by introducing a temporal notion: these texts and images were used
to evaluate the development time of a degradation in a new stone and assert the existence
of this degradation at a particular time. And if not, it was possible to estimate the onset
period. Thanks to the comparison of historical data and surveys of the present state of
degradation, a first order of size of these kinetics was estimated: the time required for a
degradation to grow up to become visually observable [4].
Fig. 2. Mappings of origin, dating and state of degradation of stones, East Tower of the castle
(Janvier-Badosa et al., 2016)
3.2 Sampling and Characterization of Degradations
The continuity of this work of visual diagnosis was laboratory characterization of

samples from the castle. The tuffeau is a calcium-silicate stone with high porosity (45 %),
containing clay minerals (10 %) [5]. The most damaging degradation for the monument
is the spalling. This degradation is characterized by a crack parallel to the surface of the
stone to 1 or 2 cm depth. Over time, a plaque several centimetres in thickness gradually
forms under the stone surface and eventually falls away, leaving the resulting new
surface powdered.
This degradation formation mechanisms are subject to assumptions not yet validated
[6]. The analyzes performed on all samples by X-ray diffraction and by ion chromatog‐
raphy showed the significant presence of gypsum (CaSO4, 2H2O exogenous mineral)
with a maximum systematically located in the cracks. Its concentration is maximum at
the fracture, approximately 1 cm depth: 1.6 weight-% of gypsum. When other networks
cracks develop parallel to the surface, greater depth the gypsum concentration increases
again locally before being zero beyond 50 mm. This gypsum distribution profile was
observed in both stones areas already engaged in a process of spalling, and in healthy
areas (Fig. 3) [7].
Fig. 3. Example of a diffractogram with gypsum detection (a), and example of the distribution
of weight-% of gypsum, depending on the depth (b).
3.3 Understanding the Spalling Process?
Artificial ageings and numerical tests were then conducted to clarify, this gypsum devel‐
opment mechanisms and to assess its role in the training process of spalling. Various
parameters were studied: the pollutant transport mode, pre-cracking of the sample,
isolation of the sample during drying. The results suggest that the test “water+gypsum”
does not allow migration of a significant amount of gypsum inside the stone.
The sulfuric acid imbibition is the method that maximizes the amount of gypsum
formed, but mainly in the 1 mm below the surface, and very little depth. Finally, the
SO2 gas followed by partial imbibition of water is the only protocol that allowed the
precipitation of a large amount of gypsum in the center of the stone, with a maximum
level on the water front (Fig. 4). The parameters linked to drying are secondary, and pre-
cracking plays only when the crack is orthogonal to the surface, allowing the sulfuric
acid to form gypsum along this crack. The distribution of gypsum, as characterized on
a flaking, seems to result from an attack by sulfuric acid (acid rain), while that charac‐
terized on a spalling appears to be from atmospheric SO2 followed by rainy periods. But
the role of the gypsum in the phenomena of scalling is not elucidated. An overall process
describing the various stages of the degradation has been proposed. This process
involves both the atmospheric SO2 and fatigue of the area subject to moisture variations
by thermal-water differential expansion. It is not possible to specify with certainty the
relative importance, or chronology. The combination of these two factors seem to be the
most probable hypothesis for the initiation and propagation of a crack parallel to the
surface, symptomatic of a spalling [8].
4 The Necessity of a Digital Tool
4.1 Specifications
Among all the data collected in this study, some data concern the building or the facade
as a whole, such as environmental or weather data. Others relate a stone or a part of the
wall, such as dating or degradation mappings. Finally, others are more specific and
concern only part of the stone, such as results of analyzes of samples. All these data, at
different scales, constitute the monument health record.
However, it was necessary to collect and organize the data in a single system,
allowing both to store and link them in order to interpret and provide heritage curators
a digital tool to better control the known data, to plan future restoration work. The
concept of digital health record takes advantage of digital technology to improve the
potential of the health record: make it more accurate, compatible with any type of docu‐
ment, easier to complete and update, analysable quantitatively, and finally usable by
many people. In this digital health, “digital” is due to 3D acquisition technologies (laser
scanner or photomodelling), structuring of data (database with 3D referencing based on
a 3D model of the monument), and the solution of work and data visualization (free and
open web interface, searchable by all professionals in heritage preservation) [9].
4.2 The 3D Model as Graphical Representation

The graphical representation should satisfy various requirements: it should be conven‐
ient to achieve, scientifically accurate and precise, and of acceptable image quality to
ensure a correct visualization of the smallest required component: a stone. So,
Fig. 4. Summary of results of aging tests.
Fig. 5. 3D model of the tower of Chaudron, with photomodeling technic (a). Selection of a 3D
entity (b) and unfolded in 2D (c).
photomodeling technique has been adopted [10, 11]. After the acquisition of geo-refer‐
enced points using a tachometer, we took pictures on a regular basis all around the tower.
The ImageModeler software has been used to calibrate and orient photographs to each
other. The building volumes were then defined as precisely as possible by the juxtapo‐
sition of geometric shapes. Textures, extracted from oriented photographs, have been
applied to the faces of the 3D model. Finally, the building was divided into coherent
architectural entities (windows, spans pilasters, capitals…) (Fig. 5).
4.3 The Database

Developed by the MAP-GAMSAU laboratory (Marseille, France), NUBES is a web-
based open source platform for the representation, documentation and analysis of archi‐
tectural features [12]. Originally, NUBES did not include the necessary features for the
creation of a digital health record on the state of conservation of buildings, and therefore
it has been adapted to match the needs of the SACRE project [13]. Once all entities are
registered, it is possible to navigate through the 3D model, and to select an entity whose
unfolded texture can be displayed in a 2D interface for semantic annotation purposes.
In order to accomplish this task, a vector graphics editor was adapted and integrated into
the NUBES platform to draw/edit mappings using hierarchical layers (Fig. 2). Compared
to a conventional mapping from a photography or drawing of a facade, all the reliefs,
so all surfaces, are mappable, including the hidden parts on an orthogonal view. Among
mapping features, a statistical analysis tool allows the automatic calculation of areas
concerning drawn surfaces (Fig. 6).
Fig. 6. The NUBES interface: drawing forms on 2D unfolded texture, displaying mappings on
the 3D model, and possibilities of statistical analysis of drawing surfaces
5 Towards a Project Application
After the SACRE project, a new project started: the MONUMENTUM project (2013–
2016). This project aims to offer curators, architects, restorers, a 3D acquisition tool,
mapping and data processing, which is both technically accessible, intuitive, and a
provider of new data that enrich the diagnosis. The SACRE project has provided the
basis for the development of MONUMENTUM project, and allowed to establish the
specifications for the implementation of the new tool. The support of these maps was
revalued, and annotations of dating, nature materials, and degradation, are made directly
on photos that allow build the model 3D photogrammetry, generating a point cloud [14].
The creation of the 3D model is thus greatly facilitated. Indeed, the technique of photo‐
modelling presented previously is very long, tedious for complex architectures and
carved parts, and sometimes cause errors during area calculations. The project also plans
to integrate the structural aspect of the building, a concept not taken into account in
NUBES and the SACRE project.
A second project, VOLUBILIS (2015–2018), also benefits progress and methodol‐
ogies developed during the SACRE project. The steps of diagnosis and the establishment
of the health record, identified during the SACRE project, are applied to the ancient site
of Volubilis.
Finally, during the SACRE project, several tourism improvements ideas have been
considered and made to present this new tool and improve the perception of the resto‐
ration by the general public in Chambord. Several web reports, regional, and newscasts
were broadcast, on the SACRE project and its implications, in order to stimulate interest
and the public’s curiosity.
Other projects were also planned: educational seminars for school, an exhibition to
present scientific results, including 3D media on computers and tablets representing
Chambord and restorations and a consultation version of the digital health record.
The issue is to convince visitors that heritage preservation is everyone’s responsi‐
bility, and that everyone can make a contribution. If these projects were not completed
during the SACRE project, these animations and exhibitions devices are now being
developed and will take place in the next months, until 2019, the anniversary of the
construction of the castle. A final project has also been created in support of this valuation
activity: VALMOD project (2014–2017). This project aims to tourism improvements
of the Castle of Chambord from the collection of architectural and historical data made
during the SACRE and MONUMENTUM projects, and their digital processing via
distribution media to different audiences. It is to work at an edutainment valuation of
knowledge gathered or produced. The public will be able to navigate through different
interpretations of the Castle of Chambord in a 3D virtual format representing different
periods to the present state and even suggestions for future states, particularly in terms
of material degradation or restorations, while accessing according to the type of audi‐
ence, in more or less popularized versions of the various sources used to offer these
interpretations.
6 Conclusions
The Castle of Chambord, emblematic monument of the Loire Valley, is the most visited
monument in the Région Centre in France. The SACRE project, providing the manager
of cultural heritage, a tool to optimize the expenses of restoration, contributed to main‐
tenance of good state of this building and ensure a sustainable tourist attraction.
The SACRE project helped drawing up the concept of the monument health record
as a tangible principle within the historical heritage conservation studies.
The three concepts of health records (past, present and future) of the monument, are
structured around a 3D graphic model, graphic support of the building and information
about it. The past of the monument is based on the history of the construction, alterations
and restorations with the various materials used. This reflects the types of degradation
observed on the castle, and extent. The future state of the monument presume the evolu‐
tion thanks to experimental and numerical studies of the mechanisms and kinetics of
degradation.
And thanks to the digital health record, quantitative statistics can be carried out on
degraded surfaces and stones: their nature, their history and their environment. The
methodology developed during the SACRE project provided the creation of new
research projects, being in line with the conclusions of the SACRE project. being in line
with the conclusions of the SACRE project.
References
1. Fitzner, B., Heinrichs, K., La Bouchardière, D.: Damage index for stone monuments. In:
Protection and Conservation of the Cultural Heritage of the Mediterranean Cities, Proceedings
of the 5th International Symposium on the Conservation of Monuments in the Mediterranean
Basin, Seville, Espagne, 5–8 April 2000, pp. 315–326 (2002)
2. Smith, B.J., Prikryl, R.: Diagnosing decay: the value of medical analogy in understanding the
weathering of building stones, building stone decay: from diagnosis to conservation. Geol.
Soc. London 271, 1–8 (2007)
3. Janvier-Badosa, S., Brunetaud, X., Beck, K., Al-Mukhtar, M.: Kinetics of stone degradation
of the Castle of Chambord – France. J. Archit. Heritage 10(1), 96–105 (2016)
4. Janvier-Badosa, S., Beck, K., Brunetaud, X., Al-Mukhtar, M.: A historical study of Chambord
castle: a basis for establishing the monument health record. Int. J. Archit. Heritage 7(3), 247–
260 (2013)
5. Beck, K., Al-Mukhtar, M., Rozenbaum, O., Rautureau, M.: Characterisation, water transfer
properties and deterioration in tuffeau: building material in the Loire valley-France. Int. J.
Build. Environ. 38(9), 1151–1162 (2003)
6. Rautureau, M.: Tendre comme la pierre, ouvrage collectif sous la direction de Michel
Rautureau, Ed. Conseil régional, Centre & University of Orléans (2001)
7. Janvier-Badosa, S., Beck, K., Brunetaud, X., Al-Mukhtar, M.: The occurrence of gypsum in
the scaling of stones at the castle of Chambord (France). J. Environ. Earth Sci. 71(11), 4751–
4759 (2014)
8. Janvier-Badosa, S., Beck, K., Brunetaud, X., Guirimand-Dufour, A., Al-Mukhtar, M.:
Gypsum and spalling decay mechanism of Tuffeau limestone. J. Environ. Earth Sci. 74(3),
2209–2221 (2015)
9. Janvier-Badosa, S., Stefani, C., Brunetaud, X., Beck, K., De Luca, L., Al-Mukhtar, M.:
Documentation and analysis of 3D mappings for monument diagnosis. In: Toniolo, L.,
Boriani, M., Guidi, G. (eds.) Built Heritage: Monitoring Conservation Management, pp. 347–
357. Springer International Publishing, Cham (2015)
10. Brunetaud, X., Stefani, C., Janvier-Badosa, S., Beck, K., Al-Mukhtar, M.: Comparison
between photomodeling and laser scanning applied to realizing 3D model for digital health
record. Eur. J. Environ. Civil Eng. 16(Supp. 1), 48–63 (2012)
11. Brunetaud, X., De Luca, L., Janvier-Badosa, S., Beck, K., Al-Mukhtar, M.: Application of
digital techniques in monument preservation. Eur. J. Environ. Civil Eng. 16(5), 543–556
(2012)
12. De Luca, L., Busayarat, C., Stefani, C., Veron, P., Florenzano, M.: A semantic-based platform
for the digital analysis of architectural heritage. Comput. Graph. 35(2), 227–241 (2011)
13. Stefani, C., Brunetaud, X., Janvier-Badosa, S., Beck, K., De Luca, L., Al-Mukhtar, M.:
Developing a toolkit for mapping and display stone alteration on a web-based documentation
platform. J. Cult. Heritage 15(1), 1–9 (2014)
14. Manuel, A., Stefani, C., De Luca, L., Veron, P.: 2D/3D semantic annotation towards a set of
spatially-oriented photographs. In: XXIV International CIPA Symposium, Strasbourg, France
(2013)
Pigments Identification Using Raman Spectroscopy
of the 16th Century Printed Book “Osorio”
Igor Lukačević1, Theodore Ganetsos2 ✉ , and Thomas Katsaros3

( )
1
Department of Physics, University J. J. Strossmayer, Trg Ljudevita Gaja 6,
31000 Osijek, Croatia
2
Laboratory of non-destructive techniques, Department of Automation,
Piraeus University of Applied Science, P. Ralli & Thivon 250, 12244 Egaleo, Greece
[email protected]
3
Byzantine and Christian Museum of Athens, Athens, Greece
Abstract. Croatia has possessed books continuously since the Middle Age. One
of the most beautiful examples of the 16th century Prandau-Normann collection
is the description of the reign of the King of Portugal Emanuel (1st) has been
written by Hieronymus Osorio − Hieronymi Osorii Lvsitani Silvensis in Algarbiis
episcope, printed in the printing house of Arnold Birckmann, one of the three
most famous Middle Age printers in Köln. Dr. Igor Lukačević, in his experimental
work [1] showed that a pigment palette is a common one for the period between
16th and 19th century. Three complementary, non-invasive spectroscopic tech‐
niques were used: micro-Raman spectroscopy, PIXE spectroscopy and UV-VIS
FORS spectroscopy. Several pigments were identified, like vermilion or cinnabar
and minium, white lead and massicot. However, pigments from blue, light blue
and green coloured regions could not be determined uniquely, leaving the authors’
palette incomplete. Fluorescence, coming from the usage of the Ar+ laser, was
the main negative factor during the Raman experiments. For some of the
pigments, it was so intense that it covered all of the pigments spectral lines. Dr.
Theodore Ganetsos, during his visit in Croatia, used a portable Raman Spec‐
trometer (laser 785 nm) [2, 3], which would not induce such fluorescence and,
consequently, more Raman lines are presented, making the pigment identification
more definite and authors’ palette complete. We identified ponsjakite to the dark
blue area, from the results of PIXE and the Raman peaks.
Keywords: Pigments · Raman · Osario · Identification · Fluorescence
1 Introduction
Lorem In the fundus of the Department of printed books in the Museum of Slavonijai
in Osijek, Croatia, the library of Prandau-Normann family stands out by its worth both
in quantity and historical significance. Almost 9000 books were brought to the Museum
from their county chateau in Valpovo by the act of the Croatian Commission for the
gathering and preservation of cultural goods (KOMZA). Their family library is an
unavoidable fund for studying the cultural history of Slavonija, and the spring of

DOI: 10.1007/978-3-319-48496-9_55
692 I. Lukačević et al.
numerous social studies, for it testifies about the fields of interest, social developments
and activities of its owners, and its content is equivalent to the contents of the collections
from other castles in Croatia and Central Europe.
One of the most beautiful examples of the 16th century Prandau-Normann collection
is the description of the military campaign of the Spanish king Emanuel Hieronymus
Osorio − Hieronymi Osorii Lvsitani Silvensis in Algarbiis episcopi…, printed in the
printing house of Arnold Birckmann, one of the three most famous Middle Age printers
in Köln. The binding of the book is covered with brown leather and decorated with what
on sight seems to be a golden oval medallion with an oriental motif. ‘‘Osorio’’ is deco‐
rated with seven foreedge miniatures, which were the centerpoint of our research: two
on the upper and lower fore-edges and three on the side fore-edge.
The sizes of miniatures are about 5 × 4 cm and their author utilized several
different colors: red, green, light blue, dark blue, yellow, pink and brown. They are
all placed on a what appears to be a gilded fore-edge. The miniatures represent (two
upper, two lower and three on the side, respectively): the four Evangelists, the Ador‐
ation of the Magi, an aristocratic coat of arms of an unknown origin and the Birth of
Jesus (Fig. 1). The inner side of the book cover is filled with the hand written notes
of, now, light brownish color.
Fig. 1. Miniatures on the upper (a), lower (b) and side of the fore-edge (c). Black circles and
labels denote the areas which were used in the experiments: A − gold, B − red, C − pink, D − dark
blue, E − light blue, F − green, G − yellow, H − brown. (Color figure online)
Pigments Identification Using Raman Spectroscopy 693
2 Experimental Results - Discussion
Fluorescence, coming from the usage of the Ar+ laser, was the main negative factor
during the Raman experiments in a previous work. For some of the pigments, it was so
large that it covered all of the pigments spectral lines. For these pigments, we could only
bring assumptions based on the PIXE results. Fluorescence has, thus, prevented us to
obtain the pigment identification with certainty for several colors. We presumed that
using the laser with the large wavelength (785 nm) would not induce such fluorescence
and, consequently, more Raman lines would be visible, making the pigment identifica‐
tion more definite.
For this reason we used the DeltaNU 785 nm portable Raman Spectrometer. (Fig. 2)
Fluorescence emission stems from sample molecules or trace impurities that absorb the
laser excitation and emit a broad background at the same energies as the Raman scat‐
tering. One way to eliminate or reduce the fluorescence emission is to select a laser
excitation wavelength that does not have enough energy to excite molecular fluores‐
cence. The advantage of 785 nm reduces competing fluorescence interference in
compounds through this process.
Fig. 2. The 785 nm DeltaNu Raman spectrometer.

It has 35 micron laser spot size, resolution 5 cm−1 and spectral range 200–2000 cm−1.
A digital microscope and a color video camera used to define the best distance between
substrate and the laser beam.
2.1 Raman Experimental Data

The 785 nm excitation appears to be the most efficient laser line. It also has the great
advantage that it can be used to perform “out of the lab” measurements with the portable
Raman set up. The recording is difficult because of a huge fluorescence, in some cases,
but not higher than that observed with the fixed instrument. A baseline subtraction is
mandatory to make the specific signatures visible. The instrument support a power of ~5–
10 mW with x50 objective, which is an important limitation for measurements with a
mobile set-up. There is a lot of ambient light collected even when a black textile is used
to cover the setup.
Gold Color. We measured several positions and especially for the gold pigment two
of them, the position A (Fig. 3) (main peaks at 1641, 1218, 755, 537 and 348 cm−1) and
the position A1 (Fig. 4) (main peaks at 1627, 1218, 813, 424 and 348 cm−1); the peaks
may correspond to the degradation of an alloy copper, zinc, gold and iron in very good
agreement with bibliography [6].
200000
180000 Gold pigment A

400000
160000 Raman peaks: 348,537,755,1218,1641 B
Gold pigment
350000
140000 Alloy Zn, Cu, Au, Fe Raman peaks: 348,424,813,1218,1637
300000
Raman Intensity
120000
Alloy Zn, Au, Cu,Fe
Raman Intensity
250000
100000
200000
80000
60000 150000
40000 100000
20000 50000
0 0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1 -1
Wavelength (cm ) Wavelength (cm )
Fig. 3. Raman spectra of position A Fig. 4. Raman spectra of position A1.
Red Color. We measured several positions and especially for the red pigment two of
them, the position B (Fig. 5) (main peaks at 336, 353, 408 and 548 cm−1) and the position
B1 (Fig. 6) (main peaks at 336, 353, 410 and 548 cm−1); in very good agreement with
the results of PIXE (Fe, Hg, Pb). We can’t specify the form of the sulfide of mercury
because the mineral and the artificial form haven’t any chemical difference. On the other
hand in PIXE we found Fe and an amount of this, is perhaps related to the mineral form
of the mercury sulfide as cinnabar. According to the Raman peaks and the PIXE results
we suggest the mixed of minium-cinnabar and red ochre. This way of application
(minium and cinnabar) is a common painting technique in medieval times.
80000
Red pigment B 50000
-1 Red pigment
70000 minium (336, 548cm ) B1
-1
vermillion (353cm ) minium (red lead) (336, 548cm-1)
60000
-1
red ochre (408cm ) 40000 -1
red ochre (410cm )
Intensity (Arb. Units)
50000
Raman Intensity
-1
40000 vermillion (353cm )
30000 30000
20000
10000 20000
-10000
400 600 800
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1 -1
Wavelength (cm ) Wavenumber (cm )
Fig. 5. Raman spectra of position B Fig. 6. Raman spectra of position B1.
Pink Color. We measured several positions and especially for the pink pigment two
of them, the position C (Fig. 7) (main peaks at 543, 1056 cm−1) and the position C1
(Fig. 8) (main peaks at 549, 1053 cm−1); the bold peaks may correspond to minium and
lead white) in very good agreement with the PIXE results.
160000
Pink pigment C
140000
120000
Minium - lead white

100000
Raman Intensity
-1
543cm
80000 -1
1056cm
60000
40000
20000
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavelength (cm )
Fig. 7. Raman spectra of position C.
120000
Pink pigment C1
100000
minium + lead white
80000 -1
549cm
Raman Intensity
60000
-1
1053cm
40000
20000
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavelength (cm )
Fig. 8. Raman spectra of position C1.
Blue Color. We measured several positions and especially for the blue pigment two of
them, the position D – dark blue pigment (Fig. 9) (main peaks at 335, 403, 746, 767,
1093, 1580 and 1623 cm−1) and the position E – light blue pigment (Fig. 10) (main peaks
at 401, 746, 769 and 1623 cm−1); the peaks (665, 667, 687 and 1050 cm−1) may corre‐
spond to Lead white) in very good agreement with bibliography. To the Fig. 9 (dark
blue pigment) we identified the peaks of posnjakite and this is in good agreement with
the previous measurements using PIXE and Raman. The last peak (1906 cm−1) Fig. 9,
corresponds to the SO4 symmetric stretch in accordance with the mineral posnjakite.
90000
80000 Dark blue pigment D
70000 Azurite Raman peaks:335,403,746,767,1093,1580,1623
60000
Raman Intensity
50000
40000
30000
20000 -1
Posnjakite (317,363,731,1033,1273, 1906cm )
10000
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavenumber (cm )
Fig. 9. Raman spectra of position D
16000
Light Blue pigment B

14000
Azurite Raman peaks:401,746, 769, 1623
12000
Lead white (665,667,687,1050)
Raman Intensity
10000
8000
6000
4000
2000
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavenumber (cm )
Fig. 10. Raman spectra of position E.
Green Color. We measured especially for the green pigment one position, the position
F (Fig. 11) (main peaks at 348, 436, 532, 755, 1054, 1086, 1232, 1498 cm−1); the most
peaks correspond to malachite) in agreement with Clark [5]. This result is in accordance
with PIXE results [1].
Yellow Color. We measured several positions and especially for the yellow pigment
three of them, the position G (Fig. 12) (main peaks at 289 and 360 cm−1), the position
G1 (Fig. 13) (main peaks at 292 and 364 cm−1) and the position G2 (Fig. 14) (main peaks
at 284 and 360 cm−1); these peaks may correspond to massicot - PbO) in agreement with
bibliography and PIXE measurements [5].
35000
Green pigment F
30000 peaks: 348, 436, 532, 755,1054, 1086, 1232, 1498
25000 Malachite
Raman Intensity
20000
15000
10000
5000
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavenumber (cm )
Fig. 11. Raman spectra of position F
200000
180000 Raman peaks: 289,360 Yellow pigment G
160000 Massicot
140000
Raman Intensity
120000
100000
80000
60000
40000
20000
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavenumb(cm )
Fig. 12. Raman spectra of position G.
Fig. 13. Raman spectra of position G1.

140000
Raman peaks: 292,364
Yellow pigment B
120000
100000 Massicot
Raman Intensity
80000
60000
40000
20000
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavenumber (cm )
Fig. 14. Raman spectra of position G2.
Brown Color. We measured several positions and especially for the brown pigment
two of them, the position H (Fig. 15) (main peaks at 551,1322,1580 cm−1) and the
position H1 (Fig. 16) (main peaks at 543, 1325, 1579 cm−1); the first peaks may corre‐
spond to the minium and the bold peaks corresponds to carbon [4].
120000
Brown pigment H
100000
minium (551) and carbon(1322, 1580)

80000
Raman Intensity
60000
40000
20000
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavelength (cm )
Fig. 15. Raman spectra of position H.
140000
Brown pigment B
120000
100000
minium (543) + carbon (1325, 1579)
Raman Intensity
80000
60000
40000
20000
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
-1
Wavelength (cm )
Fig. 16. Raman spectra of position H1.

2.2 Results - Identification

In comparison with other work we summarize our experimental results. Using the
785 nm we identify the pigments in several position to this artefact. We present our
results as shown in Table 1.
Table 1. Comparison with results from other work [1].

Area PIXE [1] Raman (514 nm) Area (this work) Raman (785 nm)
[1] (this work)
A (gilded) Gold A and A1 Alloy Zn, Cu, Au,
Pb,Fe
B (red) Vermilion and Vermilion and B and B1 Red ochre,
minium minium minium and
(HgS) vermilion
C (pink) Minium and Minium and C and C1 Minium + lead
white lead white lead white
D (dark blue) Any copper blue Any copper blue D Azurite + posnja
pigment pigment kite
E (light blue) Any copper blue Any copper blue E Azurite + Lead
pigment and pigment white
white
F (green) Malachite or F Malachite
verdigris or a CuCO3.Cu(OH)2
mixture of copper
blue pigment and
massicot
G (yellow) Massicot G, G1, G2 (PbO) Massicot
H (brown) minium H, H1 Minium + carbon
3 Conclusion
In this research work we used 785 nm Raman spectrometer in order to identify the
pigments. It was a successful procedure and we have very good agreement with related
bibliography and with other similar techniques and authors [1, 4–7].
Acknowledgements. Th. Ganetsos would like to thanks the Head of the University J. J.
Strossmayer, Department of Physics and the partners in Croatia for this useful collaboration and
the support during his stay in Osijek.
References
1. Lukačević, I., Ergotić, I., Vinaj, M.: Non-destructive analyses of 16th century printed book
“Osorio” with the colorful fore-edge miniatures. Croat. Chem. Acta 86(2), 207–214 (2013)
2. Katsaros, T., Ganetsos, T.: Raman characterization of gemstones from the collection of the
Byzantine & Christian Museum. Archaeology 1(2), 7–14 (2012)
3. Ganetsos, T., Katsaros, T., Vandenabeele, P., Greiff, S., Hartmann, S.: Raman spectroscopy as
a tool for garnet analysis and investigation on samples from different sources. Int. J. Mater.
Chem. 3(1), 5–9 (2013)
4. Colomban, P., Mancini, D.: Lacquerware pigment identification with fixed and mobile Raman
microspectrometers: a potential technique to differentiate original/fake artworks. In: Arts, vol. 2,
pp. 111–123 (2013). doi:10.3390/arts2030111
5. Bell, I.M., Clark, R.J.H., Gibbs, P.J.: Raman spectroscopic library of natural and synthetic
pigments. Christopher Ingold Laboratories, University College London, 20 Gordon Street,
London WC1H 0AJ, UK
6. Deneckerea, A., Leeflang, M., Bloemc, M., Chavannes-Mazel, C.A., Vekemans, B., Vinczed,
L., Vandenabeelee, P., Moensa, L.: The use of mobile Raman spectroscopy to compare three
full-page miniatures from the breviary of Arnold of Egmond. Spectrochim. Acta, Part A 83,
194–199 (2011)
7. Martens, W., Frost, R.L., Kloprogge, J.T., Williams, P.A.: Raman spectroscopic study of the
basic copper sukphates – implications for copper corrosions and bronze disease. J. Raman
Spectrosc. 34(2), 145–151 (2003)
Design and Application of a Data System for the
Comparative Study of Historic Mortars
Vasiliki Pachta ✉ and Ioanna Papayianni

( )
Laboratory of Building Materials, Department of Civil Engineering, Aristotle University

Thessaloniki, Thessaloniki, Greece
{vpachta,papayian}@civil.auth.gr
Abstract. Mortars are among the first building materials used in constructions
and have played a significant role in building technology’s evolution. A large
number of mortar samples were systematically analyzed, leading to the need of
a flexible data system in order to evaluate and comparatively study all results.
This system allowed recording and classifying the data input (physico-mechan‐
ical, chemical characteristics), according to the mortar type (structural, renders-
plasters, mosaic-mural substrates). With a specific toolbox all information could
be easily sorted and comparatively - statistically evaluated, while the data input
could be updated for future needs. The basic goal of the data system was to manage
the information regarding historic mortars, but throughout its use it seems that a
lot of other parameters could be also envisaged. It could therefore become a
necessary tool for any scientist engaged to the field of restoration materials and
techniques.
Keywords: Data system · Historic mortars · Statistic evaluation · Properties
1 Introduction
Mortars are among the primer materials used in construction, firstly as coverings of
wooden huts, later as bonding materials of adobe houses and stone masonries, and finally
as renders and plasters. They were often used as substrates for decorative elements such
as murals, wall or floor mosaics. Each mortar type (structural, renders-plasters,
substrates of murals, wall-floor mosaics) presents specific characteristics that are mostly
related to their functional role and application, the available raw materials, as well as
the building technology of each era [1–4].
The use of computational methods for the statistical evaluation of research results is
a well known process with many applications [5–8]. However their use for the compa‐
rative evaluation of building materials’ properties and especially of historic materials is
rather limited and usually focuses on the study of a limited number of results [9–12].
With the help of an expert system, all data of the analytical study can be easily
evaluated in giving a better profile of the analyzed mortar. The benefits of such a consid‐
eration provide:

DOI: 10.1007/978-3-319-48496-9_56
702 V. Pachta and I. Papayianni
• comparative study of great sources of information that could not be otherwise imple‐
mented
• extraction of crucial conclusions related to historic materials’ structure,
• contribute to the optimum design of the suitable repair mortar.
A further step could be the recording of similar expert systems of the characteristics of
mortars used for the repair of historic structures. By this way, comparisons could be
allowed about the characteristics and properties of existing/old/new mortars.
This paper concerns the design and application of a flexible data system for the
comparative study of a high number of historic mortars research results, analyzed during
the last two decades in the Laboratory of Building Materials of the Aristotle University
of Thessaloniki. The management of this source of information emerged many difficul‐
ties, due to the high number of samples and the large range of results. The tools that
technology provides were therefore necessary for their classification and exploitation.
By this way valuable information concerning the technological evolution of mortars was
revealed.
The study includes the description of the system’s functionality and use, as well as
indicative results from the statistical approach of historic structural mortars’ properties.
2 Design and Application of the Data System
During the last two decades more than 3000 historic mortars were analyzed in the frame
of National and European projects, according to a holistic methodology developed in
the Laboratory of Building Materials [2, 3, 13–15]. According to the methodology the
microstructural, physico-mechanical and chemical characteristics of mortars were deter‐
mined through a series of tests. Microstructure observation was performed with stereo‐
scope (Leica Wild M10) assisted by image analysis (ProgRes), aggregates granulometry
by sieving after decomposition and granulation of the original sample by hand
(EN1015-1:1998), porosity and apparent specific gravity were according to RILEM CPC
11.3, while mechanical strength was evaluated in sound, shape modified samples by
applying crushing test. Finally, wet chemical analysis assisted by Atomic Absorption
was performed in a fine fraction of the sample (<63 μm).
The comparative study of this source of information emerged many difficulties, due
to the high number of samples and the large range of results. Therefore, a logic system
of the classification of all data was of great importance. For example, which of the
characteristics of monuments, apart from their dating could be related to the mortars?
The type of the structure, such as castle, theatre? Which of the results of the chemical
results of the mortars would be considered for their characterization? Should relevant
indices be used? How the position of the mortar in the structure influences its charac‐
teristics?
The major outcome of the data system was that through the statistical analysis,
numerous conclusions could be extracted, regarding the technological characteristics of
mortars (i.e. binder/aggregate ratio, aggregates’ type and gradation, binder system,
preservation state). Remarks in relation to other parameters, such as historic period, type
of mortar, type of monument could also be made.
Design and Application of a Data System for the Comparative Study 703
The main goal of the expert system was to be easily applicable by any user, include
a great number of results, allow statistical evaluation of the data and permit modifications
according to future needs. By this way, important information for scientists dealing with
the conservation and restoration of monuments (restorers, architects, civil engineers,
conservators) could be easily drawn, since all that information could be used for the
design of compatible repair materials and for understanding the construction philosophy
of old masons.
The system was designed in Microsoft Office ACCESS 2003 and in its present form
includes the results from the analysis of totally 1500 mortar samples. It is still updated,
and it is believed that in its final form it will consist of the results of more than 3000
mortar samples.
2.1 Structure of the Data System

The data system was designed in order to upload and manage a great number of mortars’
results of various types, historic periods and monuments. Its proper design led to its
limited size (1.5 GB), so as not a central PC for its use to be necessary, but in order for
all archives to be easily transferred to any computer. It totally includes it 839 folders
and 8.069 files, from which 2.500 photos, 2.209 tables and 850 diagrams, which are all
added in the form of hyperlinks, so as not to increase its size.
The basic axes of its design were the ‘tables’ and ‘forms’ which were interrelated.
In the tables, all data was input and could be classified according to a parametric statistic
analysis, while forms were the tools for the presentation of the data [3]. The function
and usage of the data system can be graphically depicted in Fig. 1.
Data system
Tables
Forms
(Data input and
(Data presentation)
classification)
Table type II
Table type I
Monument Form type I Form type II
Historic period
(Information, photos) Front-page Selection of Historic period
Table type III Table type IV Form type III Form type IV
Mortar type -Experimental results, Selection of Monument and Experimental results of
(Structural, render-plaster, according to mortar type
mural, floor-wall mosaic of mortar type each mortar sample
-Photos, diagrams, tables
substrate, roof covering)
Fig. 1. Graphic design of the data system usage and functionality

In the data system totally 1381 mortar samples were recorded which referred to 985
structural, 300 renders-plasters, 55 floor mosaic substrates, 15 floor substrates, 18 mural
substrates, 4 wall mosaic substrates and 4 roof coverings. These samples were taken
from totally 122 monuments mainly of Greece, which were dated from the Hellenistic
up to the pre-cement modern period (beginning of the 20th century).
2.2 Data Input. Tables and Forms

The data system consisted of totally 4 types of tables (Fig. 1), which were all interrelated
[3, 9]. The first type (Table I) included the historic periods from of the mortar samples
(Hellenistic, Roman, Byzantine, Ottoman, Medieval, Modern). The second one
(Table II) gave all relevant information concerning the monuments from which the
mortars have been sampled, such as name, constructional type (castle, church, sanctuary,
bath, private house, public building), constructional period and subsequent phase etc.
(Fig. 2). Additionally, figures and architectural designs have been hyperlinked in order
to enrich all input.
Fig. 2. Table type II. Monuments’ information
The third table (Table III) included the list of the various mortar types (structural,
renders-plasters, mural substrates, roof coverings, wall mosaic substrates, floor
substrates, floor mosaic substrates). The classification of the research results according
to the mortar types was significant, since as it has been formerly concluded the selected
raw materials and the followed application techniques are diachronically interlinked to
their type [3, 13]. For example renders and plasters are usually applied in 2 to 3 well
compacted layers of reduced width, aggregates gradation and increased Binder/Aggre‐
gate ratio to the surface, which influences the physico-mechanical aspects of the layers
(porosity, mechanical strength) [15].
Table III was directly linked to the fourth table type (Table IV) which was divided
in as many tables as the mortar types, in order for the results input to be according to
the specific characteristics of each type and the samples of each category to be easily
comparatively evaluated. In Table IV (Fig. 3) the analysis results of each mortar sample
were presented (figures macro-micro were added), concerning:
• sample data (number, type, monument, historic period, sampling area)
• microstructural characteristics (type and origin of aggregates, inclusions, structure
stability, cracks and voids, salt presence etc.)
• physico-mechanical properties (porosity, apparent specific gravity, compressive
strength)
• aggregates gradation
• chemical characteristics
• other interrelated properties such as Binder/Aggregate ratio and binding system
Fig. 3. Table type IV. Experimental results of each mortar sample
On the other hand, forms were divided in four types. The first one (Form I) concerned
the front page of the data system (Fig. 4), while the second one (Form II) was interrelated
to Table type I and introduced the historic periods (Hellenistic, Roman, Byzantine,
Ottoman, Medieval, Modern) that the user could select in order to see the results of
specific mortar samples.
In the third form type (Form III), which was interrelated to Tables II and III, one
could select a specific monument and a specific mortar type and be automatically trans‐
ferred to the form type IV (interlinked to Table IV), where the analysis results of mortars
were shown (Fig. 5).
Fig. 4. Form type I. Front-page
Fig. 5. Form type IV. Analysis results from a specific mortar sample (Galerius Palace, render)
The basic advantage of the data base was that the user could sort the preferred infor‐
mation (i.e. the results from a specific mortar type of a monument) or take general
information regarding a whole category of samples (i.e. results regarding the structural
mortars of a specific historic period) and statistically evaluate them. The statistic analysis
of the results was realized through the queries that enable the user to track down and
handle specific information. These queries could be based in one or more relevant tables
and accumulate data from them.
3 Statistical Evaluation of Experimental Results
The application of the data system allowed evaluating comparatively all results, according
to the parametric classification of inputs, by grouping specific parameters [13–15]. This
parametric classification could be realized in the tables, where all results have been
uploaded. Per example, one could compare the mortars’ properties according to the mortar
types, the historic periods, specific monuments and proceed to a statistical evaluation of
results, in order to take different pieces of information.
The major goal was that through the analysis numerous conclusions could be
asserted, regarding several topics relevant to the technological evolution of historic
mortars. Indicatively, some general conclusions from the comparative study of structural
mortars are presented in this paper (binding system type, gradation of aggregates, binder/
aggregate ratio), in order to understand the data base’s functionality and effectiveness.
Figure 6, depicts the samples’ frequency according to the era to which they are dated.
Fig. 6. Frequency of mortar samples (structural, renders) according to the era to which they are
dated (total number of analyzed samples 1381)
The use of the data system, six years after its construction proved that a lot of param‐
eters could be comparatively studied, while due to the high number of results, one could
lead to safe conclusions about historic mortars’ properties. The statistical evaluation of
all results can therefore depend on the user’s interest that can configure all data.
From the statistical evaluation of the analysis results of structural mortars dated from
the Hellenistic period (4th century BC) until the beginning of the 20th century the
following observations arose [3, 13]:
The binding system type of the analyzed structural mortars throughout the historic
periods was mainly based on lime (Fig. 7). From the Hellenistic period, mixed binding
systems (lime+pozzolan) were used (8 % from the total number of Hellenistic samples
consisted of lime+pozzolan), while the majority of Hellenistic mortars (69 %) consisted
of lime+clayish material with pozzolanic reaction.
Fig. 7. Binding system of structural mortars of all eras
The use of pure lime mortars firstly presented during Roman times (21 % from the
total number of Roman samples), while the highest percentage of Roman structural
mortars consisted of lime+pozzolan. Brick dust also started to be added in the matrix
during the Roman period (5 % from the total number of Roman samples consisted of
lime+brick dust) and continued until the Byzantine and Ottoman years. In Byzantium,
mainly pure lime mortars (34 % from the total number of Byzantine samples) and lime
+pozzolan mortars were used (35 % from the total number of Byzantine samples), while
Ottoman mortars mainly consisted of pure lime (51 % from the total number of Ottoman
samples). In Medieval times the majority referred to lime+pozzolan (64 % from the total
number of medieval samples) and in Modern pre-cement period the highest number of
samples concerned pure lime (20 %) and lime+pozzolan (37 %).
Regarding the aggregates’ granulometry of the structural mortars (Fig. 8), the most
usual gradation used were 0–6 mm to 0–8 mm. Small scale alterations were envisaged,
such as:
Hellenistic mortars presented a variety in aggregates granulometry ranging from 0–
2.5 mm to 0–16 mm, but mainly consisted of aggregates of 0–4 mm size (38 % from the
total number of Hellenistic mortars). Roman mortars mainly concerned 0–8 mm size
aggregates (45 %), while 0–6 mm and 0–16 mm size were also seen (20 % and 26 %
respectively); the same was asserted for Byzantine mortars. Ottoman mortars followed
the same criteria, with mainly 0–6 mm aggregates (46 % of the total Ottoman samples),
Medieval presented a wider range from 0–1 mm to 0–16 mm and finally modern mortars
consisted of aggregates of 0–2.5 mm to 0–16 mm.
4 Conclusions
The use of computational methods for the statistical approach of research results,
regarding historic mortars seems to be a significant tool for the comparative study of
large sources of information that could not be otherwise exploited. Flexible data systems
Fig. 8. Aggregates’ granulometry of structural mortars of all eras
can therefore become important tools for the evaluation of results and the assertion of
conclusions concerning the properties and consistency of materials. Their proper design
and construction could lead to:
• archiving of a great source of information,
• presentation of all data in an easily comprehendible format,
• easy applicability by any user,
• possibility of alterations according to future needs,
• interrelation with other data systems.
In addition, these systems could act as interdisciplinary tools among scientists dealing
with the conservation of monuments, in order to disseminate knowledge and draw
valuable information regarding building materials and techniques. The collaboration
among research centers and institutions through a network of databases concerning
historic building materials, could be therefore an overall and ambitious aim, that could
lead to the evaluation of a wide range of results, such as materials characteristics, monu‐
ments types, environmental aspects, pathology symptoms etc. These networks could
become significant tools for any scientist dealing with restoration issues, while it could
contribute to the selection of proper intervention materials and techniques, for the benefit
of Cultural Heritage.
Acknowledgments. The authors would like to thank the Hellenic Cement Industry Association
for supporting the project entitled “Study of the technological evolution of mortars and concretes”
(2006-2009).
References
1. Elsen, J.: Microscopy of historic mortars - a review. Cem. Concr. Res. 36, 1416–1424 (2006)
2. Papayianni, I.: Design of compatible repair materials for the restoration of monuments. Int.
J. Restor. 10(6), 623–636 (2004)
3. Pachta, V.: Study of the technological evolution of mortars. Dissertation, Aristotle University
of Thessaloniki, Thessaloniki (2011)
4. Van Hees, R.P.J., Binda, L., Papayianni, I., Toumbakari, E.: Characterisation and damage
analysis of old mortars. Mater. Struct. 37(273), 644–648 (2004)
5. Ashlock, D.: Evolutionary Computation for Modelling and Optimization. Springer, New York
(2006)
6. Baykasoğlu, A., Güllü, H., Çanakçı, H., Özbakır, L.: Prediction of compressive and tensile
strength of limestone via genetic programming. Expert Syst. Appl. 35(1–2), 111–123 (2008)
7. Comby-Peyrot, I., Bernard, F., Bouchard, P.O., Bay, F., Garcia-Diaz, E.: Development and
validation of a 3D computational tool to describe concrete behaviour at mesoscale.
Application to the alkali-silica reaction. Comput. Mater. Sci. 46, 1163–1177 (2009)
8. Dimitriu, R.C., Bhadeshia, H.K.D.H., Fillon, C., Poloni, C.: Strength of ferritic steels: neural
networks and genetic programming. Mater. Manuf. Processes 24(1), 10–15 (2009)
9. Papayianni, I., Pachta, V., Iliadou, K.: A data base system for managing information
concerning historical mortars. In: Ioannides, M., Addison, A., Georgopoulos, A., Kalisperis
L. (eds.) VSMM 2008, pp. 271–277 (2008)
10. Moropoulou, A., Polikreti, K., Bakolas, A., Michailidis, P.: Correlation of physicochemical
and mechanical properties of historical mortars and classification by multivariate statistics.
Cem. Concr. Res. 33, 891–898 (2003)
11. Miriello, D., Barca, D., Bloise, A., Ciarallo, A., Crisci, G.M., De Rose, T., Gattuso, C.,
Gazineo, F., La Russa, M.F.: Characterisation of archaeological mortars from Pompeii
(Campania, Italy) and identification of construction phases by compositional data analysis. J.
Archaeol. Sci. 37, 2207–2223 (2010)
12. Anzani, A., Binda, L., Carpinteri, A., Invernizzi, S., Lacidogna, G.: A multilevel approach
for the damage assessment of Historic masonry towers. J. Cult. Heritage 11, 459–470 (2010)
13. Pachta, V., Stefanidou, M., Konopissi, S., Papayianni, I.: Technological evolution of historic
structural mortars. J. Civ. Eng. Architect. 8(7), 846–854 (2014)
14. Stefanidou, M., Pachta, V., Konopissi, S., Karkadelidou, F., Papayianni, I.: Analysis and
characterization of hydraulic mortars from ancient cisterns and baths in Greece. Mater. Struct.
47(4), 571–580 (2013)
15. Stefanidou, M., Pachta, V.: Technological characteristics of coatings applied in monuments
of the 4th C BC in Greece. Rev. Rom. Mater. / Rom. J. Mater. 45(1), 35–42 (2015)
GIS Applications for a New Approach to the Analysis
of Panel Paintings
Laura Baratin ✉ , Sara Bertozzi, Elvio Moretti, and Roberto Saccuman

( )
DiSPeA, Conservation and Restoration School,

University of Urbino Carlo Bo, Urbino, Italy
{laura.baratin,sara.bertozzi,elvio.moretti}@uniurb.it
Abstract. A work of art, considered in its complexity and in its evolution over
time, requires knowledge and thorough study in order to arrive at its correct
interpretation, a prerequisite for any conservation and maintenance interventions.
The evaluation of the preliminary information on the work of art and its analytical
reading are closely interrelated to a careful and critical use of the technical and
operational instruments defined in a comprehensive and focused methodological
programme. Information technology and the integration of multidisciplinary
knowledge lead to making even more powerful forms of support available for a
phase of documentation, as a means of investigation and organisation of the
information, followed by analysis and processing that implement the knowledge
of the work of art. Instruments for the 3D surveys of the panels and software for
processing and post-processing allow us to obtain digital models of the surveyed
surfaces, which are implemented in the GIS environment. We take advantage of
the advanced analytical and management capacities, normally applied to the
territory, with a simple change of scale, allowing us to carry out detailed inves‐
tigations on the painting, on the paint film and on the supporting panel. Quanti‐
tative evaluations of the metric/statistical type on the spatial distribution of the
elements are flanked both by colourimetric analysis, vectorising the RGB compo‐
nents and extrapolating the useful information, and by graphic analysis of the
iconographic composition and on the conservation status. We then process the
three-dimensional data relative to the morphology of the panel, allowing the
identification of any critical aspects or elements of deterioration, until reaching a
geometrical comparison among subsequent acquisitions that allows us to identify
any displacement due to modifications of the support. The application of the
method also allows us to define a comparison between diverse sensing systems
to verify their accuracy and effectiveness, in a perspective of programmed
management of interventions that will optimise costs and benefits and predispose
the elevation profiles to study the morphological evolution along particular lines
of interest. The GIS application in a field which is so different from the usual
context of usage provides innovative scenarios and various potentials of data
analysis and processing.
Keywords: GIS analysis · Spatial Analyst · 3D Analyst · Paintings

DOI: 10.1007/978-3-319-48496-9_57
712 L. Baratin et al.
1 Information Technology and Works of Art
A comprehensive path of evaluation and analysis of a work of art, in particular a panel

painting, provides complete and in-depth preliminary knowledge of the object and a
thorough definition of historical data connected thereto. The instruments offered today
by Information Technology lead to specific methodological paths that are well defined,
applicable with greater frequency to the entire field of the Cultural Heritage, but also
specifically to the analysis and management of painted works of art on supports such as
canvas or wooden panels. The entire process of study, documentation and in-depth study
on the work of art, on its validity and historical importance, on techniques etc., then
passes to diagnostic investigations and the surveying phase, and lastly to the relationship
between the content and its representation, which respects the metric and space char‐
acteristics. The survey “as a discipline that uses all the sciences and of all the techniques
that can contribute to the reading, to the measurement, to the analysis of the objects in
their morphological, material and structural aspects, whether overt or hidden” [1] allows
us to reach the definition of the image that is corrected, with graphic renderings and 3D
models. The image of the painting itself, if surveyed and represented with repeatable
criteria, becomes the object of a series of branches of Information and Communication
Technology defined as Image Processing, Pattern Recognition and Computer Vision [2].
In the context of restoration these concepts can take on a certain importance if you thinks
about the possibility of acting virtually on the work of art in all the operating phases,
from the preliminary study, to the evaluation on the conservation status, up to applying
various tests for a complete and correct restoration intervention. Instrumental analysis
automatically highlights and extrapolates criticalities that are barely detectable to the
naked eye with an added value of a certain importance, objectivity, integrating the
specific expertise of various disciplines applied synergistically [3]. Among the various
software applicable to the study of these works of art, of great interest and innovation
is the application of GIS systems. We speak of tools for analysis and management of
the territory that, using various types of data, can perform spatial, statistical, morpho‐
logical and morphometric analyses, identifying the characteristics that define the terri‐
tory. Applying the tool to the analysis of paintings on wood means simply changing the
dimensions of the subject under investigation. From the conceptual point of view, there
are no significant differences, since GIS operates without the scale ratio normally applied
in cartography, replacing it with the concept of accuracy, for which the absolute dimen‐
sions of the object represented are not as important as the quality with which it is
surveyed. In the case of a territory, a survey in which the metre is used as the unit of
measurement can be considered as excellent resolution, just as a tenth or a hundredth of
a millimetre is data of absolute precision in the analysis of the morphometric charac‐
teristics of a painted panel. Another element of difference only in appearance is the
georeferencing data. In a territorial context, data concerning a defined portion of the
territory are placed in “Overlay”, using the coordinates that refer to a specific geodetic
reference system, while in analysing mobile objects, data are all positioned according
to a layer of reference, however maintaining their spatial and elevation components
intact. This creates a series of digital maps of the work of art that can be represented in
the two- or three-dimensional manner, by parameterising the type and the display
GIS Applications for a New Approach to the Analysis 713
algorithm as an aid for the identification of sensitive areas, such as for problems of
deterioration or to highlight problems of deformation of the substrate which can have
an impact even on the paint film.
Analytically studying the morphology of a panel always tends to give a “synchro‐
nous” reading of its condition, that is to say, it is analysed for what it is in that moment
and not for what was previously was or will be in the future. It is therefore clear that
such an operation is always possible for the object under examination, regardless of its
evolution over time. If instead we want to have a different approach, it is necessary to
proceed to a “diachronic” analysis, that is, surveys in successive moments with the same
methods and subsequent data comparison. Precisely because a study of this kind implies
a strong interdisciplinary approach and an indispensable interoperability of data, the
maps that are obtained are always themed, calibrated and customised in order to also
make the analysis scientifically accurate, clear and understandable even to non-experts
of the GIS system.
2 GIS Methodology
The acquisition of all the data relating to the panel and the painting necessarily pass
through a painstaking phase of surveying and data processing. The creation of 3D digital
models, that accurately describe the surface of the object under examination, passes
through a process called 3D Scanning, with the creation of point clouds and the definition
of models on triangular grids to obtain a faithful representation of the shape and colour
characteristics of an actual object, for the study of errors of measurement, inconsistencies
of the shapes, accurate up to a few tenths of a millimetre. We then proceed to the
acquisition of high-resolution images, rendered along with the calibration of the camera,
map of the depth and the quality of each image, which are reworked during post-
processing and used as the basis for the creation of graphic bases and for the subsequent
phases of pictorial and colour analysis. The digital data becomes a complex and precise
set of measures in the space from which it is possible to extrapolate the significant
elements for geometric interpretation.
The passage into a GIS environment allows a very delicate phase of data acquisition,
wherein all files defined by the various stages of surveying and processing must be
standardised, translated and possibly georeferenced, to obtain a single homogeneous
project and the overlay of the information layers. We use the release of ESRI ArcGIS
10.3, proceeding in the two-dimensional ArcMap environment, for the entire phase of
organisation and data processing, to move then into ArcScene for viewing and managing
3D data. ArcCatalog allows the structuring of the apparatus of the project and the
management of the various types of layers, while the diverse instruments offered by
ArcToolbox and the various extensions and toolbars enable us to carry out a series of
in-depth analyses and data processing to obtain information and quantitative and stat‐
istical assessments.
All input data inserted in the project passes through two main stages, a first stage for
the general management of the data organisation and georeferencing of the features,
along with an organic structure of the databases, to proceed to spatial processing through
the Patch Analyst and Spatial Analyst extensions and a colourimetric analysis of the
paint film vectorising the image pixels to proceed to statistical evaluations of distribution
and frequency. The second phase focuses on the files of the rendering of the 3D reliefs,
by performing a series of operations that allow us, after organisation and harmonisation
of the data, to obtain three-dimensional processed images for the morphological and
morphometrical analysis of the panel, to make diachronic comparisons using the same
instrument or synchronic ones to evaluate the reliability of the various instruments
(Fig. 1).
Fig. 1. The GIS project is structured by means of the methodological steps defined, involving
two major phases of action structured in repeatable and standardised steps.
2.1 The Painting and the Paint Film

The organisation of the whole GIS project starts from the input process of all useful data
in ArcMap and their positioning in the same spatial context. We are not referring to a
geographic information system but one that will use the workspace as a spatial plan. The
Layer System is not georeferenced according to a geodetic reference system but instead
is only attributed units of measure. In this step, the interoperability of the software is
crucial. The descriptive layers of the graphic base and the geometric conformation of
the panel are normally created in Autocad, where it is necessary that the design is devel‐
oped in 1:1 scale, clearly defining the units of measure used. The same must be done
when the 3D relief is carried out and if the various files do not coincide as units, it will
be fundamental to scale the DWG files or the DXF 3D files (possibly created by the
WRL or by 3DS) so that everything is standardised. Obviously, it is best to always
maintain files of greater detail by scaling the rest in order not to lose the accuracy of the
best data. The units of measurement of the work plan are therefore indicated as well as
those of the display by proceeding to “georeference” everything, translating the various

files to place them in the overlay, by means of Georeferencing or Spatial Adjustment.
The procedure involves moving all the files by positioning them on a layer that defines
the X axis, approximately at the 0.0 coordinates, but the weight of the 3D files is also
an important factor to consider. Often this file is of extreme detail, and therefore, very
heavy. In this case, it is opportune to maintain the latter in the position attributed to their
insertion and shift all the other files in the overlay. These operations are often delicate,
because the three-dimensional file, once placed in the plane, loses the defined contours
of the edges and it can be difficult to interpret, for example, where the frame or support
ends. In this case, it may be useful to open these files first ArcScene in order to understand
their conformation, and then use them in ArcMap, overlaying the contour, graphic base,
orthophotos, etc. The organisation of all data prepares the project for the processing.
The graphic base or any reliefs of criticalities, such as, lacunae, loosening and detach‐
ment, abrasions, etc., define the patches with certain attributes listed in tables and as such can
be treated through the Patch Analyst extension. This is an extension that allows us to perform
spatial analysis of the coverage of a landscape, based on the conformation of the patches and
their associated attributes. The idea of comparing the patterns present on the surface of the
painting to those defining a landscape has derived from the work of Henriques and
Goncalves [4] readapted and developed for specific situations. The metrics characteristic and
conformations of the fragments that compose the painting are evaluated with statistical
indexes in relation to the attributes that define them and/or to the total area of the painting,
to highlight the evolution of the discontinuities present, quantify their distribution and make
any stylistic comparisons. The lacunae or abrasions present or any applications of gesso, if
graphically delineated, are statistically evaluated on the basis of homogeneity/heterogeneity
of distribution, regularity/irregularity of the shapes and dimensions. We then proceed to the
application of the Contour, a Spatial Analyst tool that automatically creates vector lines that
join continuous phenomena on a raster, identifying the pixels with the same value. Again,
according to a geographic concept, we are talking about an instrument of creation of the
contour lines, level curves, which allow us to identify discontinuity on the panel or on the
painted surface, due therefore also to different thicknesses of colour. The evolution of the
analysis then continues to the paint film as colour analysis. The painting is given resolution
according to the pixels that make up the image, of very high resolution, that is, defined by
their RGB values. We do not act on the entire painting because it would involve quite elab‐
orate and heavy operations for the processor, but we identify the details of attention on
which to perform the analysis. To be able to extract them, we use the vector elements that
define them, so the ideal is a good graphic base, polygonal if possible.
A series of steps enable allow us to get the details still as raster and subject them to
vectorisation processes, first for single-band to ultimately bringing everything together
in a single layer where the polygonal pixels, individual or grouped by the same value,
correspond to the same number of polygons with information relating to the three bands
in different fields of the attributes on which we can rapidly view statistical reports,
creating a bar graph of the distribution of values, indicating the values of counting, sum,
maximum, minimum, mean, median, and standard deviation. It also creates an RGB
field with the unit value, to then proceed to frequency analysis that helps us individuate
the most widespread colourimetric triads and their spatial location making it possible to
compare these data with reconstructions performed instead in the diagnostic phase,
which gives information regarding what was supposed to have been the original colour
for that detail.
2.2 Processing on Three-Dimensional Surfaces

Renderings of the 3D surveys are acquired in ArcGIS to analyse the distribution of the
data and perform calculations that allow us to rapidly individuate the morphological
evolution of the panel and any changes in diachronic investigations. The file formats,
normally VRML, are transformed into DXF and imported into ArcMap, georeferencing
all the data (if the weight of the file is excessive, for an expeditious handling we can also
directly import the *.wrl, transforming it into Multipatch and later into 3D raster to be
subjected to spatial processing, however losing some of the quality of the results). In
this phase, a very useful element is the verso of the panel, both as image and as the
vectorial rendering of the principal geometries, in order to compare any disconnection
or discontinuous movement from the support structure. For the overlay analysis these
elements are first mirrored then georeferenced, in order to superimpose on them layers
of the front as if they were being seen in transparency.
By means of the DXF we proceed directly to the creation of TIN files (Triangulated
Irregular Network). These are models that consent the representation of an area in its
three dimensions through a network of contiguous triangles, defined as nodes and lines
that are unevenly distributed, built through a technique known as Triangulation of
Delauney. For a numerical representation of the surface of the object under examination,
we use a mathematical relationship of the type z = f (x, y), whereby the surface of the
support with the paint film is described by a function that to each pair of position coor‐
dinates (x, y) associates a value in the elevation of the surface itself. The surface is then
represented through a series of triangles (TIN model) starting from a discrete number
of elevation measurements (samples consisting of x, y, z triplets) obtained from the
survey with the laser scanner, photogrammetry or photomodelling. The TIN allows us
to immediately obtain a clear visualisation of the morphology of the panel, based on the
Elevation symbology, but it also allows us to observe spatial analyses displayed by the
Symbology menu, without the need of further data processing stages, avoiding data
redundancy and taking advantage of the concept of dynamic cartography typical of the
GIS system, unique spatial information with which to associate different views based
on the same attributes of the file. The representation for Elevation highlights the height
modifications of the TIN elements through special colour scales, algorithms of definition
of the intervals and the number of classes, to then switch to the Slope and Aspect type
of analyses, which are based on the calculation of the derivative before the elevation.
The analysis of the steepness or the descent of the slopes, SLOPE, defines the slope of
the triangulations, in degrees or percentage, always with colour scales, definition of the
break interval and number of classes. The slope value for each cell is calculated as the
ratio between the height difference (rise) between the current cell and its immediate
neighbour and the distance (run) between the centres of the two cells, expressed in
degrees or as a percentage. When the input data is a TIN, the system will automatically
convert into a GRID model to obtain a homogeneous distribution of the cells over the
entire area. The slope also assumes different values depending on the direction consid‐
ered, given that the calculation, for each cell, evaluates the variation in elevation between
it and the eight adjacent cells, which are 3 × 3 floating windows, but the distance between
the centres of two adjacent cells will be different depending on the direction considered.
The slope at any surface can then assume different directions that can be grouped into
classes using the ASPECT tool. Aspect identifies the direction in the downslope of the
maximum rate of change in value from each cell compared to its neighbouring cells. It
can therefore be thought of as the display of the direction of the slope, where the values
of each output raster cell are displayed as if a compass were being used and measuring
clockwise in degrees from 0 (north or the pinnacle) to 360 (again at the northern
pinnacle), coming back to the starting point. To the flat areas that do not have a
descending direction is assigned the default value of −1, and always displayed in grey
with the indication of “flat”.
The diachronic analysis is an operation that allows the Restorer to make a qualitative
assessment of his or her intervention on an overall scale, but it is also possible to make
a quantitative assessment using the elevation profiles. In fact, it is not a given that a
deformation distributed in a more even way on the panel is not to the detriment of the
increase of the curvature. To evaluate this specific problem, it is possible automatically
derive the application of elevation profiles from templates starting from TIN or DEM
modes, thanks to a specific function of the 3D Analyst for ArcGIS. The “Interpolate
Line” tool leads to the definition of a shape file that follows the morphological trend of
the model along a line of defined length that can be reapplied to various TINs in overlay,
allowing timely comparisons. Customisable graphics are created with the elevation
profile, which can be subjected to exaggeration to better highlight the discontinuities.
3 Results
The Spatial Statistics menu of the Patch Analyst allows us to define a series of indexes
on the panel of the Madonna and Child with Saint Monica, Saint Augustine and the
Trinity by Vittore Belliniano, both for the analysis of the application of gesso carried
out on the lacunae, and of the iconographic base allows us to define how the artist may
have worked in the context of a stylistic comparison and/or of the activity of the
“bottega” (workshop) (Fig. 2). The first 4 indexes, SDI (Shannon’s Diversity Index),
SDE (Shannon’s Evenness Index), AWMSI (Area Weighted Mean Shape Index) and
MSI (Mean Shape Index), apply only to the landscape level and evaluate the number,
form and complexity of the patches. We see in this case, for example, an SDI of 0.203,
an SEI of 0.2936 and an MSI of 1.4666. The SDI is equal to 0 if there is only one patch
and grows with the increase of the number, the SEI is equal to 0 when the patch distri‐
bution observed is low and approaches 1 when the distribution of types of patches
becomes more uniform and the MSI is equal to 1 when all patches are circular and
increases with the increasing of the irregularity of the shape. Already it is possible to
identify a certain presence of patches, a rather heterogeneous distribution and a certain
unevenness of the shapes. In the analysis for the classes of the application of gesso on
the total area, it detects the presence of 4037 patches (NumP), an average perimeter per
patch of 0.0329 (MPE), an average size of 0.000042 (MPS) and total areas of all the
patches defining the application of gesso amounting to 0.17108 on a total of 3.30832 m.
Then there are the indexes indicating the median, standard deviation, coefficient of
variation of the patches, etc.
Fig. 2. The Patch Analyst extension allows us to apply a series of indexes for statistical
evaluations of both the criticalities present on the panel and of the conformation of the patches
that make up the iconographical base.
The automation of the recognition of the values helps us to rapidly identify discon‐
tinuities, assess the depth according to the proximity of the contour lines and define the
contours objectively. In Fig. 3 the Contour application on the panel by Lorenzo Lotto.
Vectorialization of pixel also offers the possibility of managing the three fields sepa‐
rately allows us to assume a digital restoration, carrying out restoration tests on the screen
with the inclusion of the values that the finished intervention should have and evaluating
the quality of the rendering without having to act directly on the painting. It reproduces
a virtual reality of the painting, in which the machine and not the human eye, identifies
the colours that are present. The vectorial pixels, original or restored, can be associated
with the correct symbology as correspondence of colour through the attribution of a
specific script.
In Fig. 4, for example, a detail of the Carlo Crivelli Altarpiece shows that the triad
of greatest distribution in the bodice examined is 75.71.73, while diagnostic investiga‐
tions have revealed that the original colour should have corresponded to an ultramarine
blue, therefore 18.10.143.
An interesting case study examined with this method is the “Hermit Saints Triptych”
by Hieronymus Bosch where the panel was surveyed with different techniques, laser
scanning, photogrammetry and photomodelling, and in three different campaigns which
allowed us to carry out an analytical study of a diachronic type.
Fig. 3. The Contour tool of the Spatial Analyst leads to the creation of contour lines which bring
together the elements of equal value recognised by the raster with the orthophotos of the painting,
enabling us to both already highlight the demarcation of the junction between the wooden boards
(as you can clearly see from the layer of the verso placed in overlay), and to delineate contours
of highlighted details or of criticality (stains, lacunae…).
In Fig. 5 there is a comparison between the three successive campaigns, the first
made on 4/10/2013, the second on 12/12/2013 and the third on 18/11/2015 at the
conclusion of the restoration, using a Faro laser. Between the first and second interven‐
tions, there had been an intervention on the panel involving the removal of the wood
block structure and the result already shows evident differences, clearly identifiable from
the processing carried out. The left edge of the panel in the second campaign appears
lower and quite uniform from the bottom to the top. The upper right corner is still the
lowest area but it is certainly less imbalanced compared to the first campaign. The right
edge of the panel also looks more uniform and continuous. Considered overall, the
morphological trend in the second campaign is more homogeneous and regular. The
data obtained with the third campaign tend to confirm the anomalies already detected
but the panel appears with a more uniform morphology, perhaps related with the cleaning
done on the paint film. Therefore, the diachronic analysis confirms that there has been
some adjustment of the panel although the anomalies on the top part still persist, lower
than the bottom section, while the imbalance of the left side seems to have been reduced,
at least in the lower part. The analysis of the data obtained using the Aspect tool of
ArcGIS, in Fig. 5 at the bottom, viewing with the same settings values through 8 classes
plus the Flat one, leads to a number of conclusions:
1. The panel as a whole has a general deformation that extends along the diagonal
direction from the upper left corner to the lower right corner.
2. Still very evident is the disconnection located along the entire strip of about 3 cm
positioned at the top of the panel.
3. The diachronic analysis shows a much more homogeneous pictorial surface in the
third campaign compared to the first two, which can be related to a successful
cleaning intervention. All this of course makes it possible to positively assess the
operations implemented during the intervention and demonstrates how the proposed
methodology is able to record the operations that have been carried out.
Fig. 4. The detail of the bodice of one of the figures of the Crivelli Altarpiece panel is vectored
to perform statistical and distribution analyses of the RGB values and any digital restoration
operations by comparing the actual values with the original ones identified through diagnostic
investigations. (Color figure online)
4. The right panel of the support is slightly more irregular than the one on the left, but
overall the support appears homogeneous, with both parts facing outwards, creating
a panel that is curved and slightly warped.
The Slope indicates that the database contains the values that also refer to the thickness
of the panel and not only of the painted surface, which, dimensionally much larger, tend
to blend together all the differences found on the surface in question. Notice how the
red colouring is arranged exclusively on the edges. All this, however, does not prevent
us from confirming the anomalies due to warping and those due to the 3 cm strip on the
top part of the object. The third campaign highlights anomalies that are not easily
explainable with diachronic analysis.
The interpretation of the processing is also supported by a comparison with the image
and the graphic base of the verso of the panel, mirrored and placed in overlay. In this
case it has been useful for making clear how the reversal of direction does not take place
at the junction of the panels in the verso, located in the perfect centre about 30 cm from
Fig. 5. Viewing of the elevation data of the three campaigns conducted with the Faro laser scanner
on the central panel of the Hermit Saints Triptych by Hieronymus Bosch. The 3D surveys have
been subject to processing that have enabled the creation of a TIN from which to obtain data on
elevation (at their peak), on the slopes (at their centre, Slope) and on their direction (ASPECT, at
their base) to make diachronic comparisons between the three successive campaigns.
the edge, but is shifted by about 3 cm to the left of the centre, which can easily be
explained by the thinning it underwent in 1838.
In Fig. 6 are shown for comparison the profiles derived from the first two campaigns
carried out with laser, in this case Minolta, to highlight the comparison. The interpolation
lines are positioned in the lower part of the panel, the part that presents the maximum
convexity, having a progression from left to right and naturally they have been obtained
always along the same track. The comparison shows that the highest part of the panel,
corresponding to the central zone, has not undergone significant differences, while
recording a clear lowering of the external areas already on the right, but even more
apparent on the left and this of course entails a greater total convexity. The scale on the
axes X and Y is in mm.
Fig. 6. The automatic creation of elevation profiles along lines of interpolation defined on the
surface of the panel help identify the morphological evolution to highlight significant differences.
4 Conclusions
A completely innovative methodology both in the context of GIS systems and in the
applications to the cultural heritage is presented, which has led to the development of a
system of research which attempts to bring together the various needs and technical-
theoretical competence to develop a system that is repeatable and standardised. The
possibility of analysing the panels in their diverse artistic-historical-structural aspects
that make up their composition allows us to delineate an overall picture of the work of
art in its current state to define intervention operations.
Certainly fundamental in the development of a research of this entity has been the
possibility of working on various highly prestigious works of art, such as the triptych
of the Martyrdom of St Liberata and the Hermit Saints Triptych by Hieronymus Bosch,
undergoing restoration by the Gallery of the Accademia of Venice and by the Dutch
Bosch Foundation, the Polyptych Altarpiece by Carlo Crivelli of Montefiore dell’Aso
(Ascoli Piceno), the Madonna and Child in Glory with the Saints Martin, Michael the
Archangel and George by Vincenzo Pagani of the Civic Museum of Ripatransone
(Ascoli Piceno) and the Madonna and Child with St Augustine, Saint Monica, All Saints
and the Trinity by Vittore Belliniano, of the Civic Museum of Feltre (Belluno).
To make diachronic comparisons between surveys carried out with the same meth‐
odology in different periods also directs attention to any deformations and transforma‐
tions of the panel while it is possible to use the processing carried out also for comparing
synchronic surveys, but made with different instruments, such as laser scanners, photo‐
grammetry or photomodelling, evaluating the reliability of the data to assess the best
practices in terms of cost/accuracy.
References
1. Baratin, L., Bertozzi, S., Moretti, E., Saccuman, R.: Monitoring of the deformation of the
support with laser technologies and analysis of the pictorial surface with GIS systems. In:
Proceedings of APLAR 5, Vatican City, Rome, 18–19 September 2014 (2014)
2. Bertozzi, S., Baratin, L., Moretti, E.: Pictorial surfaces and supports: GIS analysis,
characterisation and monitoring. In: Proceedings of ESRI Italia 2015, Ergife Palace Hotel,
Rome, 15–16 April 2015 (2015)
3. Bennardi, D., Furferi, R.: Il Restauro Virtuale tra ideologia e metodologia. Edifir (Edizioni
Firenze S.r.l.), Florence (2007)
4. Henriques, F., Gonçalves, A.: Analysis of lacunae and retouching areas in panel paintings using
landscape metrics. In: Ioannides, M., Fellner, D., Georgopoulos, A., Hadjimitsis, D.G. (eds.)
EuroMed 2010. LNCS, vol. 6436, pp. 99–109. Springer, Heidelberg (2010). doi:
10.1007/978-3-642-16873-4_8
Project Paper: Visualisation, VR and AR
Methods and Applications
‘Translation’ and Fruition of an Ancient Book Through
Virtual Reality in the Case of Lost Cultural Heritage
Case Study: “Inscriptiones” by Emmanuel Thesaurus
Sanaz Davardoust1 ✉ , Anna Osello1, and Rosa Tamborrino2

( )
1
Department of Structural, Geotechnical and Building Engineering,
Polytechnic of Turin, Turin, Italy
{sanaz.davardoust,anna.osello}@polito.it
2
Department of Architecture and Design, Polytechnic of Turin, Turin, Italy
[email protected]
Abstract. This article presents a methodology for historical information fruition,

such as ancient books texts, in the tourism field. The case study highlights the
great possibility that virtual reality (VR) offers to Cultural Heritage professionals
in terms of communication and awareness raising of end users. This methodology,
applied to the rare 17th century book by Emmanuel Thesaurus “Inscriptiones
quotquot reperiri potuerunt Opera ed diligentia Emmanuelis Philiberti
Panealbi”, allows us to show the lost seventeenth-century traits of the Marble
Hall in Palazzo di Città (Turin). Tourists can be immersed in a virtual space based
on the words of the ancient book within the real space of the Hall, which let him
relive the early baroque project atmosphere.
Through this work, it is possible to show how the use of cutting-edge ICT,
such as BIM, can impact both on research and society, by arousing the public
interest for Cultural Heritage and shared understanding.
Keywords: Digital history · Virtual reality · Information fruition · Historical

building · Digital/virtual documentation · Digitized book · Protection ·
Valorization · Interoperability · BIM
1 Introduction
The subject of the historical and artistic heritage protection is currently the focus of
many debates since it concerns a field having one of the greatest economic potential,
especially in Italy: the tourism. [10] Italy possesses a very rich and interesting multi‐
millenial history, providing the community of a wide range Cultural Heritage, consisting
of building, spaces, paintings, books and an almost unlimited source of information and
written and physical documentation.
Nevertheless this abundance poses many management, degradation prevention and
effective communication problems while ensuring economic sustainability investments
in this field. Moreover, in most cases, the information is guarded in barely accessible
and readable manuscripts and rare books. The original language is often Latin, which

DOI: 10.1007/978-3-319-48496-9_58
728 S. Davardoust et al.
was kept in use in some fields until late 20th century, for example in homily and religious
inscriptions, but is not used anymore except from an elite inner circle of scholars and
intellectual (People usually can’t understand Latin although some Italians continue to
study it at school). Besides, many ancient and original books are preserved in historical
libraries, whose access and fruition is often limited, particularly for non-professionals
users. Digitization started to bridge this gap allowing the access also to rare books and
documents difficult to reach (i.e. the ongoing digitization of the Archivio di Stato in
Venice by a project of Ecole Polytechnique de Lausanne [12], or the realization of Search
engines and databases, such as “data.bnf.fr.” or “Europeana”, which inform about where
to find different prints and libraries around the world.
Another important aspect concerns the communication of such information, which
mostly are just readable words, which need interpretation by highly specialized scholars
and professionals. In fact, this information can be enigmatic for most of the people, since
they were directed to an audience of other epochs and with specific interpretative capa‐
bilities based on profound knowledge of a historical, social, artistic, and political context.
Nevertheless, these documents often include much more than simple notions for disci‐
plines: they are witness of a space and a time of civilization, becoming especially
important for lost Cultural Heritage.
This state of art makes necessary the figure of experienced professionals with specific
skills in reading, interpreting, filtering and processing information, in order to show the
data to end users in the most effective way.
In this sense, the extremely rapid and stable development of ICT is constantly
providing innovative solutions both to facilitate research operations and to elaborate
new prospects of development and protection of the Cultural Heritage, especially when
Data interpretation need to be also visualized. Most books in fact were not illustrated
and now they are perceived with an incomplete use of the sense of the view.
In recent years, for example, the Virtual Reality is increasingly imposing thanks to
its ability to communicate tangible (physical and geometrical elements) and intangible
information (historical data, energy, structural, etc.) through a three-dimensional recon‐
struction of the descriptions taken by historical sources. VR is the artificial reconstruc‐
tion of both real and imagined - or no longer existing [17, 18] - environment with the
support of software and devices for three-dimensional modeling and user/model inter‐
action. This innovative technology uses cutting-edge digitization and visualization tools
to immediately and effectively show tangible and intangible aspects.
Many projects, since late 90 s and still in our days, highlighted the possibilities of
VR models in the field of tourism, dissemination and sensitization of common users;
among the most advanced it’s possible to cite the virtual tours of Casa Batllò (Barcelona)
[15], Valle del Tevere and Teramo [16]. But the implementation of ICT in CH could go
beyond the simple 3d restitution, and help the communities in Historic Estate manage‐
ment. BIM methodology is an important example of how to integrate powerful database
management, interoperability with other software and complete support to technical
users involved in a historical building preservation, while providing highly communi‐
cative 3d models and data for VR environment modelling.
For this reason, BIM and VR can now work translating information coming from
ancient and cryptic sources into highly understandable and immersive ambient made of
‘Translation’ and Fruition of an Ancient Book 729
colors, shapes and sounds. The resulting visualization of such works is usually entrusted
to various kinds of computer devices, from the traditional monitors of computer plat‐
forms to the most modern immersive navigation systems, such as Oculus Rift.
The treated case study in this article, Inscriptiones by the Count Emmanuel
Thesaurus, is a collection of the paintings descriptions once belonging to Palazzo di
Città, Municipality of the City of Turin since 1659.
2 The Aula Maior of Palazzo di Città as Described

in “Inscriptiones” by Emmanuel Thesaurus
The Inscriptiones quotquot reperiri potuerunt opera et diligentia Emmanuelis

Philiberti Panealbi of the Jesuite Emanuel Thesaurus o Tesauro (Turin 1593–1677),
were published in Turin in 1666. Among the several editions of the books in libraries
around the world, there is its copy in the National Library in Turin, Biblioteca Reale.
This issue is very rare and precious, and its contents is very relevant for the history of
the city and its lost Cultural Heritage (CH).
This book, after the restoration of the Town Hall in 19th century, is the last direct
source that historian can examine to find evidence of the original baroque appearance
of the once called Aula Maior. In fact, the building went through constant expansions
and transformations until 20th century and often was at the center of dramatic events of
the city.
The sieges who threatened the city, the economic crisis due to the numerous wars of
the XVII e XIX centuries and finally the Napoleonic army’s entrance in the capital of
the Savoy State had affected a negative effect on the public coffers, often precluding the
possibility of maintaining the original structure of buildings and urban image (See
Fig. 1).
Fig. 1. Town hall and its square, the relationship with piazza castello in a satellite view.
In some rooms of the Town Hall, the degradation reached advanced stages and forced
the Municipality to radical interventions, which did not always correspond to a simple
restoration, but rather, in some cases, led to deeper transformation.
The Marble Hall–the current name of Aula Maior - is the most significant example
in this sense, because the roof infiltration irremediably disfigured the sumptuous ceiling
paintings, made by the Flemish painter Jean Miel. The majority of the surfaces were
Fig. 2. Data collection and modeling of the marble hall
irreparably damaged, suggesting the whole restoration of the room in a neoclassical

design, which reflected the most used style of early nineteenth century [2].
The intervention completely transfigured the Baroque style and its characteristics,
replacing the wooden finishing and the ancient frescoes with the marble finishing (from
which the hall takes its actual name) and wooden ceilings decorated with gray pigmen‐
tation to simulate marble. The austere and dignified character of the hall where citizens
can meet the Mayor has finally buried the seventeenth-century image of the Hall hiding
precious and antique tales. In fact, as Cravero recalls in the first modern study published
on the history of this building, “(…) the magnificent frescoes cycle (…), illustrated the
most striking episodes of the ancient history of Turin” - “(…)l’imponente ciclo d’affre‐
schi illustravano gli episodi più salienti dell’antica storia di Torino.”- and he goes on to
say, “fortunately, a precise and direct testimony of the lost paintings was conserved by
(…) Count Emanuele Filiberto Thesaurus, - “Per fortuna delle pitture scomparse ci è
stata serbata una precisa testimonianza, (…) del Conte Emanuele Filiberto Tesauro” -
who probably inspired the choice of the themes and the Latin mottos, which, very
didactically, described the depicted scenes [4].
Thesaurus was an important figure of the period, who received the entrustment of
the frescoes themes choice for many rooms of the Municipality. He was a “Vir nobilis‐
simus, in Aula Taurinensi celeberrimus” since the 18th Century Storia Letteraria
d’Italia by Zaccaria and became the tutor of the sons of the Carignano Duke, being a
very reputed erudite by the Savoy aristocracy. He wrote in Latin, but he also translated
from the Latin adapting ancient texts. This book, which is now preserved in the Royal
Library of Turin, and other important historiographical sources, give us other important
hints on the disposition of those paintings in the hall. [6].
Some of Thesaurus’ writings have been reprinted by Doglio [7]. In this context we
aim to underline how the fruition of the original book and its related information can be
accessed by a wide range of readers for enhancing the understanding of CH. Information
can be found in libraries and museums, also by using important new digital sources. For
example, through Europeana, European Digital Library, we can search the portrait of
the author, Thesaurus, and we can find it in the Austrian National Library. [13, 14]
Nowadays we can link Thesaurus’s iconography to its biography through the web and
databases, such as French National Library dbs, which contains some of his books. [9]
With this research we aim to connect written description with the lost pictures in order
to visualize - within the space of the room - how visitors at that time had to feel the
previous designed baroque vision of this painted space. The room was a visual narrative,
and we translated the book in a rich narrative by integrating words and space, and icon‐
ography.
Cravero himself quoted Eugenio Olivero –author- hypothetic scheme, which was
the basis for the reconstruction of the frescoes disposition.
These lost pictorial narrations are also important for a profound comprehension of the
path to nobility of the City of Turin since they visualized the mythological origins of the
Savoy family. As many other Italian lordship previously made during the Renaissance, the
Savoy family, even if in belated epoch, used mythological tales of its own born as a propa‐
ganda tool in an European field, to magnify its own dignity and right to reign.
From this direct source, a meticulous reconstruction of the image of the hall begins.
This research, therefore, has a double purpose: on one hand, telling how the room
appeared before the nineteenth-century intervention, highlighting the profound stylistic
transformation and the resulting change in the visual impact on its visitor, compensating
the lack of iconographic documents with the manuscript description. On the other hand,
bringing back to tourist and visitors an insight into past Turin culture and history, valor‐
izing an important but neglected and hidden written source.
Palazzo di Città and Aula Maior have been modelled in the context of the project
“Digitization of the Buildings of Turin” which aims to realize a virtuous and innovative
model for managing Public Estate through BIM. This methodology allows the integra‐
tion of both tangible and intangible information (such as Thesaurus’ book descriptions)
inside a 3d information model; these data, which are now conserved in expensive and
hardly accessible papery archives, can be thus dematerialized, becoming available for
many typologies of users. Moreover, this source will contribute to the enrichment of a
digital web supporting the actors involved in the building and its historic heritage
conservation.
This goal is achieved through progressively structured activities: building survey,
historic, management, architectonic and plant data retrieving, 3d and information model‐
ling of the whole building in BIM environment and, finally, use of dedicated software
for specific activities such as energy analysis, Facility Management and, as deepened in
the next paragraphs, 3d immersive VR modelling for tourism, dissemination and sensi‐
tization activities.
3 Interpretation and Translation of the Book into a 3dimensional

Virtual Space
After a long process of careful reading and interpretation of the information retrieved
from the book and the other written sources, it was foreseen the possibility to translate
the words of the manuscript into a better communicating device.
Among the different opportunities provided by ICT developers, the virtual reality
has been chosen for its economic sustainability, rapidity of realization, great perceptive
impact and power of suggestion, and high usability by any end-users.
The first step was the definition of narrative character for the Hall modeling. Unde‐
niably, the lack of iconographic documentation of the lost paintings discouraged the
project from taking a photorealistic approach; the peculiarity of these scenes, and the
consequent scarce diffusion of these subjects in the paintings by contemporary artists to
Miel, indeed does not permit to identify similar paintings that would allow a plausible
and philological reconstruction of the real images. Therefore, the focus is put on the
evocative appearance of the project and its cultural and political message, limiting the
choice to preserved seventeenth-century paintings, contemporary and with similar char‐
acteristics to those of Miel, conserved in Turin museums. For this reason, in the virtual
model, the paintings are deliberately blurred, while the descriptions in the book by
Thesaurus, being reliable and first-hand, were directly re-used and accurately shown.
Nevertheless, the carried out research on the ancient system of the room and their
comparison with the Latin text, provided a number of information at least on the distri‐
bution of the images. The erudite person, in fact, gives precise description on both
subjects of the paintings, and their subdivision: the top 5 histories described the origins
of Turin. At the four corners were immortalized four episodes where history and myth
were blended, from the Roman period while at the center we could admire the Egyptian
myth of Phaeton founding the city. The believing in this legend built a great interest in
Egyptian civilization and brought a humongous amount of archeological findings to
Savoyard capital. This immense collection is now conserved in one of the most admired
and visited attraction of the city: the Egyptian Museum [8].
A description of other 10 boards on the upper strip of the wall, and 8 in the lower,
follows. The first ones carry ten further significant historical moments in the evolution
of Turin, while the latter illustrate eight emperors who, for different reasons, were
important for the city; in this case, the figures were conventionally represented between
columns, like in other examples of the time (image of Emperors between columns) [3].
The division of the strips of the wall most likely remained the same also in the later
nineteenth century restoration, which sets a sharp color change between the marbles
below the access doors transom and those above it.
Some of the hypotheses about a more precise disposition of the paintings, made by
successive historical sources, were rejected after a comparison between baroque envi‐
ronments and other seventeenth-century rooms of the palace, which survived until the
present day.
The methodology of this work involves the restitution of the certain information
through a realistic modeling, while all the uncertain data were translated into evocative
images. For example, all measures (width, height, length of the room and of the
Fig. 3. Covering the neoclassical surfaces and showing the baroque style
windows) of the virtual environment are identical to those of the real room. In fact the
basis of the 3d virtual model is the actual Marble Hall while the baroque style is visual‐
ized as a painted veil, covering the neoclassical surfaces and showing the ancient
appearance. (See Fig. 3) The starting sources were the Historical Archives projects, the
historical documents that tell the story of the Hall, computer material delivered by the
Town Hall of Turin, consisting of a.dwg file of the existing and security plan, and finally,
a rich bibliography about the building, its square and generally the history of Turin.
This information was verified and compared with the metric and photographic survey
of existing building, allowing very detailed modeling of the framework up to a 1:1 scale.
After updating.dwg, it was possible to begin the modeling phase with Revit 2014, BIM
software whose tools like schedules, parametric families, information on materials or
else, enable the implementation of an enriched three-dimensional model with easily
generated, exportable database, always consultable by the different actors who are
involved in the project [1] (See Fig. 2).
4 From BIM to VR for the Visualization of Data
Another important feature that was added to the 3d model is the possibility of reading
the words of the book from an animated pop up screen, which show the original pages
of the manuscript, preventing its degradation. To do this, the Aula Maior was specifically
characterized for the environment visualization in the virtual reality, taking advantage
of the interoperability of BIM software, and exporting the model to 3ds Max, which can
manage in a fast and efficient way, the rendering and navigation operations of the envi‐
ronment. Subsequently the file was exported to Unity, software used for gaming to create
environments where characters move. This gaming platform allows managing graphic
renderings, light effects, creating soil, physical simulation, sound effects and options for
customizing of the project (scripting systems). For example, through this last procedure,
it is possible to set an automatic pop up of the information relating to the paintings and
to objects when the user moves closer to them [5, 11].
5 Results
The result of such work is a suggestive and evocative environment that provides an
overall idea of space to the visitors by an immersion in the room.
The project brings the paintings descriptions back into their contest, the room; so
that the users can visualize the content of the book in the same ambient they were born
from (the Marble Room). In such way it is possible to recreate the lost ancient appearance
and fruition of the space, building with VR a good visual narrative in which virtual space,
text and images combine.
Moreover, the tourist can make a comparison between the Marble Hall and the Aula
Maior, moving in the real room and visualizing the ancient look through the Oculus or
other VR devices. In addition, the user can select any wanted information, personalizing
his experience and enjoying the simple visual involvement or deepen the knowledge
about Turin myths and history. The possibility of interaction with the modeled elements
in this virtual environment is one of the most important strengths of VR, because any
object can contain extra information selected and provided by designers about the epoch,
the protagonists of the histories once depicted in the paintings and the historical context
in which Thesaurus moved (Fig. 4).
Fig. 4. Displaying of the marble hall using unity
6 Conclusion
In conclusion, this research highlighted some important aspects for any other work
involving Digital History, emphasizing the innovative approach to both buildings and
books valorization. The ICT tools can in fact be a great ally in the communication of
any kind of source belonging to Cultural Heritage, thanks to their power of involving
tourists and non-professionals into highly immersive experiences, stimulating curiosity
and interest for artistic and historical richness. These technologies can be a support to
overcome any difficulty that both professionals and visitors might face in reading the
sources of first-hand information, such us manuscript, building inscriptions and any
other written data. The digitalization of our CH libraries and archives can undoubtedly
be one of the main fields of research for both scholars and professionals, providing a
solution to the conservation and managing of incalculable numbers of public heritage
treasures. Nevertheless, the project shows how the informatics technology can’t substi‐
tute the traditional phase of reading, filtering and interpretation of the historical sources,
emphasizing the important role of human factor in the crucial choices of the right tools
for the final communication of the data. The role of the designer is also determinant in
finding a reading key assisting the end-user in concentrating on the hidden details,
finding the history laying in all those environment and buildings we take for granted.
References
1. Osello, A. (ed.): Buiding Information Modelling: Geografic Information System, Augmenred

Reality for Facility Management. Dario Flaccovio, Palermo (2015)
References from books
2. AA.VV.: Il Palazzo di Città a Torino, The City Palace in Turin, Vincenzo Bona S.p.A., Turin
(1986) (in Italian)
3. Thesauri, E.: Inscriptiones quotquot reperiri potuerunt Opera ed diligentia Emmanuelis
Philibeti Panealbi, editio secunda, Taurini (1666)
4. Cravero, D.G.: Trecento anni di vita Del Palazzo Civico di Torino 1663–1963, Three hundred
years of City Palace life in Turin 1663–1963. Città di Torino, Turin (1964). (in Italian)
5. Sherman, W.S., Craig, A.B.: Understanding Virtual Reality-Interface Application and Design.
Morgan Kaufmann, San Francisco (2003)
6. Zaccaria, A.: Storia letteraria d’Italia, literary history of Italy, pp. 690–691, vol. III. Poletti,
Venice (1750) (in Italian)
7. Luisa, DM., (ed.): Emanuele Tesauro, Scritti, Emanuele Tesauro, Writing. Edizioni dell’Orso,
Alessandria (2004) (in Italian)
8. Cibrario, L.: Storia di Torino, history of Turin. Alessandro Fontana, Turin (1846). (in Italian)
9. Tamborrino R., Rinaudo F.: Translating urban history, research and sources, into interactive
digital libraries. In: Geomatics Workbooks, no 12, pp. 175–183. Como (2015)
References from other literature
10. MiBAC, 2006/2008. Ministero per i Beni e le attività culturali, CulturaItalia. http://
www.culturaitalia.it
References from websites: (Accessed 18 May 2016)
11. http://www.realtavirtuale.net/oculus-rift/
12. http://dhlab.epfl.ch/
13. http://data.bnf.fr/atelier/12182716/emanuele_tesauro/
14. http://www.europeana.eu/portal/record/92062/BibliographicResource_1000126200534.html
15. http://mobileworldcapital.com/508/
16. http://www.itabc.cnr.it/pagine/progetti-ricerca-itabc-cnr
17. http://www.raiscuola.rai.it/articoli/la-domus-aurea/21345/default.aspx
18. http://journalofdigitalhumanities.org/3-1/transforming-the-object-of-our-study-by-john-n-wall/
An Interdisciplinary Study on the Ancient Egyptian Wines:
The Egywine Project
Maria Rosa Guasch-Jané ✉

( )
Mondes Pharaoniques (UMR 8167 “Orient et Méditerranée”),

Université Paris-Sorbonne, Paris, France
[email protected]
Abstract. This article presents the research results of the ‘Irep en Kemet’ Project
that studies the Ancient Egyptian wine culture and the newly developed website
of the research project [www.wineofancientegypt.com] to transfer the knowledge
and disseminate the results. For the first time, the corpus of the viticulture and
winemaking scenes in the ancient Egyptian private tombs has been developed,
together with the bibliographical and scene-detail databases. The second phase
of the ‘Irep en Kemet’ website includes an interactive archaeological map of
Egypt with the viticulture and winemaking scenes, and also the databases and the
results of the research. Moreover, the objectives and preliminary results of the
EGYWINE project that investigates the wine jars and wine inscriptions, and the
ancient DNA of the Egyptian wines, are presented.
Keywords: Ancient egypt · Grapes · Wines · Wine jars · Inscriptions ·

Paleogenomics
1 Introduction
The research on the subject of Ancient Egyptian wines is an interdisciplinary research

including several studies and research projects.
The ‘Study of the viticulture and oenology scenes in the Egyptian tombs’ scientific
project (2011–2014) is a scientific project, directed by Dr. Maria Rosa Guasch Jané, was
hosted by the Faculdade de Ciências Sociais e Humanas of the Universidade Nova in
Lisbon (Portugal), and funded by the Fundação para a Ciência e a Tecnologia (FCT)
of the Portuguese Ministry of Education and Science. The research project has created
the corpus of the viticulture and winemaking scenes in the ancient Egyptian tombs.
The “Ancient Egypt’s Wine Rebirth” Egywine project (2016-2018) is the Marie
Sklodowska Curie Grant number 699858 (H2020-MSCA-IF-2015) funded by the Euro‐
pean Comission, and hosted by the Mondes Pharaoniques (UMR 8167 “Orient et Médi‐
térranée”) directed by Professor Pierre Tallet at the Université Paris-Sorbonne in Paris,
France. Egywine project studies the ancient Egyptian wine jars and wine inscriptions
and is carrying out the first ever paleogenomic analysis of the ancient DNA (aDNA) of
Egyptian wines. Egywine addresses the viticulture origins, production and preservation
of the ancient Egyptian wines, and the diffusion of the Egyptian wine culture legacy to

DOI: 10.1007/978-3-319-48496-9_59
738 M.R. Guasch-Jané
Europe. To transfer the knowledge and disseminate the research results, the website
‘Irep en Kemet’ [www.wineofancientegypt.com] has been developed and is now
updated with the recent research on this subject.
2 Study of the Colour of the Egyptian Wines
The symbolism of the wine in ancient Egypt was based on its red color, not only because
of the relation established between wine and the blood of the resurrection god Osiris,
but also because of the reddish color of the River Nile during the annual flood-with
ferruginous sediments coming from the Ethiopian mountains-, when the harvest time
for grapes started [1, 2]. The Egyptian mythology related the wine only to the red color,
and no textual references to white wine–or to red wine- from the Dynastic period (3100–
343 BC) have been found in Egypt up to now. The first mention of white wine in Egypt
is from Athenaeus of Naucratis, who lived during the 3rd century BC, in his book The
Deipnosophistae, where he explains that Mareotis wine, in the area of the lake Mariut
near the city of Alexandria, was “excellent, white and enjoyable, aromatic…” [3].
During my PhD at the Nutrition and Food Science Department, Pharmacy Faculty
of the Universitat de Barcelona, the kind (color) of the wines that were made in ancient
Egypt was investigated. An analytical method [4, 5] for archaeological residues of wine
was developed using the liquid chromatography mass spectrometry in tandem (LC/MS/
MS) technique. Two compounds were identified in archaeological residue samples from
Tutankhamun’s amphorae using the LC/MS/MS method: tartaric acid, as grape marker,
and syringic acid derived from malvidin, the latter being the main compound responsible
for the red color of grapes and wines, as red grape marker [4, 6, 7]. The results of
analysing residue samples from Tutankhamun’s amphorae revealed that in ancient Egypt
red and white wines were given the name irp [5, 7].
The analytical results added new information to the inscription on the amphorae:
about the type of wine contained. There was a red wine in Tutankhamun’s amphora
number Journal d’Entrée (JE) 62314 [5, 7] at the Egyptian Museum in Cairo, with the
inscription “Year 9, wine of the Estate of Aten of the Western River, chief vintner Khaa”,
and a white wine was contained in Tutankhamun’s amphora JE 62316 [5, 7] with the
inscription “Year 5, wine of the Estate of Tutankhamun, ruler of Thebes in the Western
River, chief vintner Khaa”. The results of the analyses also confirmed that in Egypt,
during the New Kingdom period, three kinds of grape products were made [5]: red wine,
white wine and the shedeh, a red wine with a different preparation.
The origin and nature of the shedeh, which has no translation, was a mystery since
a century ago, with pomegranates or grapes having been proposed as a raw material.
According to Papyrus Salt 825 (BM 10051) of the Late period (715–332 BC), the only
text found so far that mentions the elaboration of the shedeh, it was filtered and heated;
nevertheless, due to a damage in the Salt papyrus, the botanical source of shedeh
remained unknown [5, 8–10].
The results of analyzing a sample of a residue from the shedeh amphora found in
Tutankhamun’s burial chamber (JE 62315, Carter number 206 [11]), bearing the inscrip‐
tion “Year 5, shedeh of very good quality of the Estate of Aten of the Western River,
An Interdisciplinary Study on the Ancient Egyptian Wines 739
chief vintner Rer” [12] confirmed that shedeh was a red grape wine [5, 6]. Further
research suggested the use of the three different wines found in Tutankhamun’s Burial
chamber were for the King’s three-step resurrection ritual [13].
3 The Corpus of the Viticulture and Winemaking Scenes
Viticulture and winemaking scenes were depicted on the walls of the private tombs in
Egypt from the Old Kingdom period (2575–2150 BC) through the Graeco-Roman times
(332 BC–395 AD).
The ‘Study of the viticulture and oenology scenes in the Egyptian tombs’ scientific
project (2011–2014) has documented and studied the viticulture and winemaking scenes
in the Egyptian private tombs [14, 15].
For the first time, the scenes have been documented and studied together, including
the texts associated to the images, and photographed. The scenes depicted on the walls
of the tombs represent the main steps of the harvest and the wine production and are
unique (see Fig. 1).
Fig. 1. Grape harvest scene from the tomb of Khnumhotep III (BH3) at Beni Hassan, Egypt.
©‘Irep en Kemet’ Project 2016
A bibliographical researchable database (Fig. 2) relevant to wine, viticulture and

winemaking in the ancient Egypt has been completed, with 200 entries including articles,
books, chapters in book, academic thesis, essays, abstracts, on-line articles and websites.
Furthermore, a scenes-detailed database with 92 records of tombs dated from the 4th
Dynasty (2575–2450 BC) of the Old Kingdom until the 30th–31st Dynasties (380–332
BC) of the Late period has been developed [15].
Fig. 2. Bibliographic database on the Ancient Egyptian wine. ©‘Irep en Kemet’ Project 2016
The database of the scenes of viticulture and winemaking in the ancient Egyptian
private tombs (Fig. 3) includes for each scene: record number, scene description with
the different steps according to the theme, that is, viticulture or winemaking.
In the viticulture scenes, the steps represented are: vinery, taking care of the vine,
grape harvest and counting the baskets.
In the winemaking scenes, the steps represented are: transporting grapes to a press,
pressing grapes, heating and filtering, pressing the remains in a sack press, filling wine
jars, fermentation, offerings to goddess Renenutet, wine tasting, sealing the wine jars,
labelling wine jars, counting wine jars, transporting wine jars to a cellar, refrigerate
during fermentation, and store wine jars in a cellar.
The scene-detail database (Fig. 3) also includes: scene details, annotations, scene
type and scene condition, text (inscriptions and translation) and image/photo. Moreover,
the dating (period, dynasties and kings), location (provenance, governorate, archaeo‐
logical site), provenance (governorate, archaeological site, tomb name and number, and
location inside the tomb) and present location (if it is the same as provenance, if it’s a
museum then location inside the museum and inventory number, and others if it’s not a
museum) are detailed.
The project methodology included data collection analysis of the archives, compi‐
lation of documents and all the information associated with the Egyptian tombs having
scenes of viticulture and winemaking, such as location, tomb owner names and titles,
dating, among others [15].
Fig. 3. Scenes-detailed database of the Ancient Egyptian tombs with viticulture and winemaking
scenes. ©‘Irep en Kemet’ Project 2016
This comprehensive study of all scenes and associated texts, combined with a photo‐
graphic survey in Egypt during 2013, have permitted to record and analyze all the scenes,
including some unpublished ones.
The bibliographical research of the scenes of viticulture and winemaking has resulted
in the identification of unpublished scenes. Furthermore, some of the scenes of viticul‐
ture and winemaking are not available for publication, because of copyrights or because
they were never drawn or photographed. For this reason, and with permission of the
Permanent Committee and Foreign Mission Affairs Department of the Supreme Council
of Antiquities (SCA) of Egypt, we carried out a photographic survey for the tombs.
The 2013 photography mission to the Egyptian necropolises included the identified
tombs to check the information obtained from historical documents and to take updated
photographs of the scenes. The photographic mission allowed us to have access to almost
all the tombs with scenes of viticulture and winemaking to study the images and asso‐
ciated text, and to identify unpublished data. It also permitted us to detect the current
state of conservation of the already published scenes.
The photography mission work was divided into three phases, based on the
geographic location (Upper Egypt, Middle Egypt and Lower Egypt) allowed taking
detailed, high-resolution photos of the viticulture and winemaking scenes, which will
be presented in the scenes-detailed database [15].
Furthermore, our successful photography mission also gave a great opportunity to
explore the tombs, in which these scenes were depicted and learn about their contexts
and current conditions.
This research study on the viticulture and winemaking scenes in the ancient Egyptian
private tombs is an important tool for the future documentation and conservation of the
archaeological heritage of Egypt.
4 Website “Irep En Kemet”, Wine of Ancient Egypt
The website ‘Irep en Kemet’ (=Wine of Ancient Egypt) project [www.wineofancien‐

tegypt.com] is being developed (Fig. 4) to transfer the knowledge and disseminate the
results of our research on the subject of the wine culture from Ancient Egypt to the
scientific and non-scientific community.
Fig. 4. The ‘Irep en Kemet’ (=Wine of Ancient Egypt) website on the study of the Ancient
Egyptian wine culture. ©‘Irep en Kemet’ Project 2016
The creation of the website in 2013 was financially supported by the VINSEUM,
Museu de les Cultures del Vi de Catalunya, in Vilafranca del Penedès (Barcelona, Cata‐
lonia) [www.vinseum.cat] and developed by Sistemes de Gestió de Patrimoni (SGP)
[www.sgponline.net].
The website content includes the general information about the research on the viti‐
culture and winemaking scenes and including: introduction, history, objectives and
methodology, the team members, bibliography (Fig. 5), news (Fig. 6), links, collabora‐
tions and contacts.
Fig. 5. Published articles and congresses in the website. ©‘Irep en Kemet’ Project 2016
Fig. 6. Photographic mission, lectures, articles and congresses presented in the website news.
In 2015 a second stage of the website, funded by Perelada Fine Estates in Perelada
(Girona, Catalonia) [www.perelada.com/en], is developed to include the project results,
the bibliography and scenes-detailed databases, and the interactive archaeological map
of Egypt for the tombs with viticulture and winemaking scenes (Figs. 7 and 9).
Fig. 7. Archaeological map of Egypt with the tombs having viticulture and winemaking scenes
on the ‘Irep en Kemet’ website. ©‘Irep en Kemet’ Project 2016
Fig. 8. Database of the tombs having viticulture and winemaking scenes presented in the website.
Furthermore, the website includes our publications on the topic of wine in Ancient
Egypt, freely accessible on-line.
The website will include all the collected data, studies and analysis. Currently, the
on-line access to the scene-detail and bibliographic databases, and to the georeferenced
tombs’ map is being developed (Fig. 8).
Fig. 9. The viticulture and winemaking scenes presented in the website. ©‘Irep en Kemet’ Project
2016
Regarding the technical aspects, the website is multiplatform and available for
consultation through mobile devices: tablets and 3G mobiles. The Qr codes have been
generated to allow the consultation through these devices. The platform language is
based on XML (eXtensive Markub Language). The web interface is designed using
Content Management System (CMS) with open code, a free software under General
Public Licence, allowing an intuitive navigation and actualization of the contents of the
website.
Fig. 10. Tutankhamun’s amphora Journal d’Entrée number 62303. ©Maria Rosa Guasch Jané,
with permission of the Egyptian Museum in Cairo
The web platform gives visibility to our research on the Ancient Egyptian wine
culture, an open access to the scenes-detailed and bibliography databases, which will be
georeferenciated, and the dissemination and socialization of the archaeological knowl‐
edge in relation to wine in antiquity as well as bibliography and multimedia material
(images, audiovisual) and 3D reconstructions. The use of the Information and Commu‐
nications Technology (ICT) will allow dissemination and on-line consultation.
5 Egywine Project
The Egywine project studies how the Egyptian wines were made to understand the
ancient traditional elaboration methods for the knowledge advancement and heritage
conservation.
Egywine is an interdisciplinary project that collects and documents evidences
(pottery and organic material) for the entire process of grape cultivation and wine
production in Egypt interrelating various scientific disciplines (archaeology, paleoge‐
nomics, history and semantics).
The EGYWINE project aims are to:
• Identify what is linked with wine from the Predynastic (3800–3300 BC) to the New
Kingdom (1550–1069 BC) period;
• Study the Ancient Egyptian wine jars;
• Analyze ancient wine residues from pharaonic Egypt through paleogenomics.
5.1 Egyptian Wine Jars and Wine Inscriptions
Since the Predynastic Period (4000–3100 BC), wine jars were placed in the Egyptian
tombs as funerary offerings. During the New Kingdom Period (1539–1075 BC), two-
handled wine jars (amphora type) were inscribed in hieratic to indicate: the year, the
name of the product, the quality, the provenance, the property (royal or private) and the
name and title of the wine-maker (Fig. 10).
The inscription on the amphora JE 62303 (Fig. 10) of the Cairo Egyptian Museum,
from Tutankhamun’s tomb (KV62) at Thebes, reads: “Year 4, wine from the Estate of
Aten, in the Western River, chief vintner Nen” [12]. The clay seal of this amphora is
broken and it is open on the upper part, as Fig. 10 shows. Like the labels in the modern
wine bottles, the inscriptions on the New Kingdom wine jars give us information about
the harvest and wine production [16] revealing that the ancient Egyptians considered
this information relevant and necessary to be able to distinguish between wines. It was
extremely important to know the vintage and the provenance of the product.
The method of wine production and preservation of the ancient Egyptian wines, and
the diffusion of the Egyptian wine culture legacy to Europe is investigated through the
analysis of the pottery and the wine inscriptions.
5.2 Paleogenomic Analysis

The paleogenomic analysis aim to identify possible relations with current grape vari‐
eties, wine yeasts, and grape vine origins.
Egywine project offer multiple perspectives of study: (i) vine genome evolution since
its initial domestication; (ii) fermentation yeasts evolution; (iii) wine making process
evolution, through a metagenomic analysis of the multiple microorganisms used during,
or along, the wine making process.
Acknowledgements. The ‘Study of the viticulture and oenology scenes in the Egyptian tombs’
scientific project (2011–2014) was funded by the Fundação para a Ciência e a Tecnologia (FCT)
of the Portuguese Ministry of Education and Science.
The “Ancient Egypt’s Wine Rebirth” Egywine research project (2016–2018) number 699858
is a Marie Sklodowska Curie grant financially supported by the European Comission (H202-
MSCA-IF-2015). We would like to thank the Vinseum and the Perelada Fine Estates for the
financial support of the website.
References
1. Poo, M-Ch.: Wine and wine offering in the religion of ancient Egypt. In: Helck, W.,
Westendorf, W. (eds.), pp. 149–151. Kegan Paul International, London (1995)
2. Poo, M-Ch.: Weinopfer. In: Helck, W., Westendorf, W. (eds.) Lexikon der Ägyptologie, vol.
6, pp. 1186–1190. Harrassowitz, Wiesbaden (1986)
3. Athenaeus (I, 33 d-f): Athenaeus, the deipnosophists, with an English translation by Charles
Burton Gulick. In: Loeb Classical library, vol. 1, pp. 146–147. London Harvard University
Press, Cambridge-Massachusetts (1961)
4. Guasch-Jané, M.R., Ibern-Gómez, M., Andrés-Lacueva, C., Jáuregui, O., Lamuela-Raventós,
R.M.: Liquid chromatography mass spectrometry in tandem mode applied for the
identification of wine markers in residues from ancient Egyptian vessels. Anal. Chem. 76,
1672–1677 (2004)
5. Guasch-Jané, M.R.: Wine in ancient Egypt: a cultural and analytical study. In: British
Archaeological Reports S1851, pp. 11, 29–30, 53–54, 56–59. Archaeopress, Oxford (2008)
6. Guasch-Jané, M.R., Andrés-Lacueva, C., Jáuregui, O., Lamuela-Raventós, R.M.: The origin
of the ancient Egyptian drink shedeh revealed using LC/MS/MS. J. Archaeol. Sci. 33, 98–101
(2006)
7. Guasch-Jané, M.R., Andrés-Lacueva, C., Jáuregui, O., Lamuela-Raventós, R.M.: First
evidence of white wine in ancient Egypt from Tutankhamun’s tomb. J. Archaeol. Sci. 33,
1075–1080 (2006)
8. Derchain, P.: Le papyrus Salt 825 (B.M. 10051), Rituel pour la Conservation de la Vie en
Égypte. Mémoires vol. LVIII: 137 (II, 1), (no. 10), pp. 137, 147–149. Académie Royale de
Belgique, Bruxelles (1965)
9. Tallet, P.: Le shedeh, etude d’un procédé de vinification en Égypte ancienne. BIFAO 95, 459–
492 (1995)
10. Tallet, P.: De l’ivresse au dégrisement, à propos d’un article récent sur le vin shedeh. Göttinger
Miszellen 227, 105–112 (2010)
11. Carter Archive: Tutankhamun: anatomy of an excavation, The Howard Carter Archives. The
Griffith Institute, Oxford. http://www.griffith.ox.ac.uk. Accessed 15 June 2016
12. Černy, J.: Wine jars. In: Baines, J. (ed.) Hieratic Inscriptions from the Tomb of Tutankhamun,
Tutankhamun’s Tomb Series, vol. II, pp. 1–4. Griffith Institute, Oxford (1965)
13. Guasch-Jané, M.R.: The meaning of wine in Egyptian tombs: the three amphorae in
Tutankhamun’s burial chamber. Antiquity 85(329), 851–858 (2011)
14. Guasch-Jané, M.R., Fonseca, S., Ibrahim, M.: The ‘Irep en Kemet’ project, documenting the
corpus of wine in ancient Egypt. In: Proceedings of 4th International Conference (Euromed
2012) on Cultural Heritage and Digital Libraries, Amathus (Cyprus), Oct 29-Nov 3 2012,
International Journal of Heritage in the Digital Era Suppl 1, 1, pp. 181–186. ISBN 2047–4970,
Multi-Science Publisher, UK (2012)
15. Guasch-Jané, M.R., Fonseca, S., Ibrahim, M.: The complete corpus of viticulture and
winemaking scenes from the ancient Egyptian private tombs. In: Proceedings of the XIth
International Congress of Egyptologists, Florence, Italy, 23–30 August 2015 (in revision)
16. Guasch-Jané, M.R.: On Egyptian wine marketing. In: Hudecz, A., Petrik, M., (eds.)
Commerce and Economy in Ancient Egypt, Proceedings of the 3rd International Congress for
Young Egyptologists 25–27 September 2009, Budapest. BAR S2131, pp. 63–69.
Archaeopress, Oxford (2010)
ArchaeoInside: Multimodal Visualization of Augmented
Reality and Interaction with Archaeological Artifacts
Kadar Manuella and Domsa Ovidiu ✉

( )
Department of Computer Science and Engineering, 1 Decembrie 1918 University of Alba Iulia,
Alba Iulia, Romania
Abstract. This paper reports on a system named ArchaeoInside designed in

order to offer a virtual environment for archaeological exploration with large
access to public, researchers and museum curators. ArchaeoInside project aims
at recording, classifying, digitizing, accessing and presenting archaeological sites
and artifacts in Augmented Reality.
Keywords: Augmented reality · 3D interaction · VR systems · Archaeological

artifacts
1 Background
Virtual Reality (VR) and 3D modeling projects involving archaeological data started at
the end of the eighties, the beginning of the nineties. First attempts into establishing
standards and methodologies for the use of VR and 3D modeling techniques in archae‐
ology, particularly data communication by means of such techniques, the historical
credibility of VR and 3D products, the accuracy of the products from the archaeological
point of view or the need for a transparency of data have been widely discussed in [1–4].
Other methodological and theoretical aspects of VR and 3D modeling applications
in archaeology have been developed [5] and new approaches of archaeological research
using VR and 3D models have been proposed in [6–8]. Issues such as: should archae‐
ologist bother at all with VR and 3D modeling, what is the gain of the archaeological
research from VR or why should archaeologist deal with VR have received pertinent
answers in well-documented research works that emphasized the need of multi-disci‐
plinary collaboration. In the last five years, projects that integrate cultural contents,
interfaces and technological devices have been launched by interdisciplinary teams, in
order to have a holistic vision of information. The emphasis is given to perception,
capacity of learning, 3D behaviors, connectivity, dynamic processes of learning,
embodiment as fundamental factors of virtual communication.
Innovative multimodal visualization technology has opened the possibility of
creating augmented reality that integrates 3D content into a museums website, thus
enhancing the experience of learning acquired by a visitor’s interaction with an online
exhibition, either within the museum or in the Internet [9, 10]. Further, virtual reality
interfaces offer curators new technological tools for preservation and access. The curator

DOI: 10.1007/978-3-319-48496-9_60
750 K. Manuella and D. Ovidiu
may utilize such tools to extend the already existing digital preservation techniques by
adding digital 3D models of artifacts to their digital archives, and then reuse these digital
surrogates for presentations in visualization systems that also allow online access [10].
This paper reports on a system, named ArchaeoInside which has been designed by
an interdisciplinary team of archaeologists, computer graphics specialists, programmers
who have collaborated to offer a virtual environment to the large public and to
researchers and curators.
ArchaeoInside has been designed by taking into consideration key issues when
designing museum interactive system. Key issues as have been presented in [10], are:
1. Museum interactive systems should be as cost effective as possible given the limited
funds available to museums.
2. 3D content should be created as cheaply as possible in addition to digitization of
supporting data.
3. Consideration should be given to the costs of maintaining the museum interactive
systems because this implies new skills that need to be acquired by the museums,
staff etc. The museum may in effect be converting itself from a learning institute to
a so called hybrid institution where the institution exhibits are not only analogue (i.e.
the physical artifact), but also the digital surrogate or resources. In this context, it is
important that authoring tools contain all tools necessary for proper digital curation.
4. Appropriate interaction techniques should be devised to augment the digital resource
so as to effectively engage the user. In order to identify suitable interaction techni‐
ques for the end-user but also to the curator, formal usability evaluation studies are
necessary. Therefore, relevant skills are needed.
5. Any museum interactive system should present the information as a story that rein‐
forces the heritage behind the artifact that is on display targeting at different users
and age groups. Using the new opportunities that digital storytelling [11] offers it is
required to extend the skill set of museum curators and staff.
6. Perceived ‘presence’ is shown to be enhanced when modalities such as sound and
3D content are added in a museum interactive system in order for the visitor to feel
part of the virtual exhibition.
2 ArchaeoInside Model
ArchaeoInside deals with recording, classification, digitization, access and presentation

of elements that belong to archaeological excavations. The way the data are organized
and accessed follows the territorial standards logic and it is directly connected with its
contents. It is a concrete -not theoretical- open and evolving model.
The objectives of the system are: (i) documentation of the identity and value of the
prehistoric sites, (ii) gathering and preservation of historic evidence that are scattered
in public and private archives, (iii) conveying of findings and data to the local public,
competent authorities, student population through museums and to the international
academic community through web integration, (iv) development of a digital platform
for recording and documentation of the cultural heritage.
ArchaeoInside: Multimodal Visualization of Augmented Reality and Interaction 751
The system was designed to manage various types of data such as: temporal data
(description of historical periods), description of spatial data (description of places of
the archaeological site), different kind of maps that have been digitized (axonometries,
maps, sections, plans, elevations, excavation profiles and plans), digital photos or ancient
photos digitized, videos, scanned drawings, scanned texts, vectorial plans (generated in
SVG), 3D models. The Use Cases are detailed in UML schema in Fig. 1. Figure 1
presents various operations for an efficient exploitation of the ArchaeoInside system.
All UML schemas help the user (visitors and experts) to query interfaces by keywords
and images.
Fig. 1. Use Cases of ArchaeoInside
3 ArchaeoInside System
The user has the option to explore the virtual environment starting from a 3D map that
is chosen from the main menu of the application, further one can obtain information on
the region selected and can continue to ask for information on objects discovered in the
archaeological site (Fig. 2).
Fig. 2. Selection menu

The example provided here is the Seusa-Gorgan site from Alba County representing
an archaeological complex site dated to the Late Chalcolithic. The site is a settlement
belonging to Cotofeni Culture. General Information is provided on dialogue base. The
user can explore the now-a-days site (Fig. 3) through a virtual panoramic view and then
can immerse into the excavation, namely the chalcolithic village. In the prehistoric scene
the user has the option of exploring the 3D location of the chalcolithic site.
Fig. 3. Panoramic view of the site –Seusa Gorgan
If the user is more interested in the artefacts than the location, these can be accessed
directly from the menu. The artefacts are categorized by large classes such as the material
and then in more specific categories such: type of object, usage, culture, size. New
categories or subcategories can be added easily by modifying the xml. definition file
(Fig. 4).
Fig. 4. The Seusa-Gorgan site excavation
Interaction is provided through sound, interactive conversation, video presentations

or interaction with artifacts situated in the location. There are several detail zones within
the location that provide accurate information on different types of materials and made
techniques belonging to the referred period.
Visualization and manipulation of 3D objects are facilities integrated into the Object
Viewer. The user can rotate, move, and resize selected objects. Several levels of detail
are provided for textures and for the 3D models of the artifacts that can be changed by
accessing the graphical interface within the Object Viewer (Fig. 5). New objects can be
loaded without recalling another state of launch. The user has the option to select the
background gradient, object’s details and type of visualization camera.
Fig. 5. Digital bone chisel with surface details in Object Viewer
4 Methodological and Technological Approach
The challenges in generating virtual worlds are connected to performances versus

natural behavior and aspect of the environment. Such requirements are often in contra‐
diction: convincing models and high level physical simulation implies better hardware
and software, meaning an increased computational level that has effects on the global
performance. The ArchaeoInside system has been achieved by using open source or
shareware software such as: Irrlicht Engine [12], irrKlang Audio Library [13], Newton
Game Dynamics Physics Engine [14], LibCurl Multiprotocol File Transfer Library [15],
Ffmpeg Audio/Video conversion tool [16]. Using the Irrlicht Engine and irrKlang,
ArchaeoInside integrates event analysis and contextual media in its 3D interfaces.
In the engine selection process several factors have been considered: speed, stability
and usability. The Newton Game Dynamics engine was chosen due to its popularity and
because has strong community development, is fast, and has a simple API. Newton
Game Dynamics is an integrated solution for real time simulation of physical environ‐
ments. The API provides scene management, collision detection, and dynamic behavior
and yet it is small, fast, stable and easy to use. The engine implements a deterministic
solver, which is not based on traditional Link Control Protocol (LCP) or iterative
methods, but possesses the stability and speed of both, respectively. This feature makes
the product a tool not only for games, but also for any real-time physical simulation. In
the current version of the application only the collision component is used. The appli‐
cation calculates in real-time collisions between two objects. First is the static object
which represents the current loaded location and a dynamic object which simulates the
user controlled avatar. For each object specific materials can be set which describe the
interaction parameter. In future versions the interaction between more objects can be
simulated, such as the interaction with artefacts, so the user will be able to physically
interact with them besides other interaction modes.
For data acquisition, 3D modeling of the artifacts, image processing, UV mapping,
texture mapping various software has been used such as: Autodesk Maya 2012 [17],
Adobe Photoshop CS 3 Extended [18].
Museum artifacts can be digitized accurately using 3D scanners, photogrammetry
and cheap software that creates photo realistic 3D models for online display. Innovative
interaction systems have been designed to expand the traditional museum approach of
displaying an artifact in a show case with curators’ point of view on a simple card. The
idea is to liberate the physical artifact by a digital surrogate and interact with it. It has
been considered the ability to create interaction systems composed of replicas of a
Fig. 6. Display of bone chisels findings from Seusa-Gorgan

physical artifact linked to a 3D model (digital surrogate) of that artifact in order to deliver
a contextual heritage view on the artifact [10].
The show case (Fig. 6) can be considered an example for integrating Augmented
Reality (AR) through computer-generated virtual information into the real physical
environment [19]. One can imagine such a system in a museum whereby the actual
artifact is displayed in its show case, perhaps on a wall (Fig. 6), and a large display is
situated next to the case. Further, a robust physical replica of the artifact is linked to the
display, which presents a virtual environment containing a 3D model of the artifact.
Thus, the visitor will obtain tactile information that is traditionally impossible, and by
selecting attached sensors on the replica they can also explore a 3D world that digitally
narrates the story of the artifact on the display [10].
Figure 7 presents a Chalcolithic bone chisel which was imported into Maya devel‐
opment environment for model building.
Fig. 7. Chalcolitic bone chisel
Fig. 8. Bone chisel wireframe
Many developers find this arrangement to be a much better way to work because one
can ultimately create a more precise model. Attempting to model from memory or by
referring to hard copy references will ultimately lead to errors, and the proportions of
the final models will likely be off. A polygonal model has been used due to its mathe‐
matical simplicity: by providing a piecewise linear approximation to shape, polygonal
meshes lend themselves to simple, regular rendering algorithms in which the visibility
and colors of most pixels are determined by interpolating across the polygon’s surface.
Such algorithms embed well in hardware, which has in turn led to widely available
polygon rendering accelerators for every platform. In addition, polygons serve as a sort
of lowest common denominator for computer models, since most model representations
(spline, implicit-surface, volumetric iso-surface) may be converted with arbitrary
accuracy to a polygonal mesh. For these and other reasons, polygonal models are the
most common representation for every application from video games to visualization
of medical, scientific, and CAD data sets [20].
In order to import a 3D model into an engine like Irrlicht the model has to be made
up of polygons (Fig. 8). A primary concern when modeling 3D models for applications
is the polygon count and efficiency. Models have to be more efficient to maintain a
manageable data set to render. The reasoning here is that an efficient streamlined envi‐
ronment composed of the lower poly assets will render more smoothly and give better
frame to frame renders. The modeling methods applied can generate also low polygon
resolution version of the meshes without modifying the perception of the model [20].
The ArchaeoInside project is still in progress but the work done up to now already allows
us to draw methodological conclusions for the creation of a virtual reality system based
on a unique model with high accuracy and detailed data. The entire reconstruction of
the virtual site is next to be achieved. Further improvements to the system will be the
addition of input devices such as virtual reality gloves and extending the system so that
it can be used with mobile devices. Through the 3D reconstruction of the Calcholitic
village we approach an accurate and suggestive interpretation that involves the public
and transmits the importance of cultural heritage and thus its conservation.
Such models allow to disseminate data and information that often belong to the
scientific community. The possibility to interact with cultural heritage increases the
respect and the understanding of the public. The use of advanced applications and freely
shared technologies gives the possibility to make the common heritage accessible to
everyone in its historical, artistic and scientific dimensions.
Acknowledgements. Special thanks are granted to Mr. Mihai Gligor, the principal investigator
of the Seusa-Gorgan archaeological site for providing access and valuable information and to the
enthusiastic team of students hearing Computer Science, namely Robert Iancu and Vlad Buda
who have volunteered in gathering data from site, have widely contributed to the programming
and digital representations of artefacts.
References
1. Ryan, N.S.: Computer-based visualization of the Past: technical realism and historical
credibility. In: Main, P., Higgins, T., Lang, J. (eds.) Imaging the Past: Electronic Imaging and
Computer Graphics in Museums and Archaeology, No. 114 in British Museum Occasional
Papers, pp. 95–108 (1996)
2. Kanter, J.: Realism vs. reality: creating virtual reconstructions of prehistoric architecture. In:
Barceló, J.A., Forte, M., Sanders, D.H. (eds.) Virtual Reality in Archaeology. BAR
International Series, vol. 843, pp. 47–52. Archaeopress, Oxford (2000)
3. Frischer, B., Abernathy, D., Favro, D., Liverani, P., De Blaauw, S.: Virtual reality and ancient
rome: the UCLA cultural VR labs Santa Maria Maggiore project. In: Barceló, J.A., Forte, M.,
Sanders, D.H. (eds.) Virtual Reality in Archaeology. BAR International Series, vol. 843, pp.
155–162. Archaeopress, Oxford (2000)
4. Forte, M.: About virtual archaeology: disorders, cognitive interactions and virtuality. In:
Barceló, J.A., Forte, M., Sanders, D.H. (eds.) Virtual Reality in Archaeology. BAR
International Series, vol. 843, pp. 247–259. Archaeopress, Oxford (2000)
5. Vanaten, I.: Argumentation paths in information infrastructure of the archaeological virtual
realities. In: Burenhult, G. (ed.) Archaeological Informatics: Pushing the Envelope CAA 2001.
BAR International Series, vol. 1016, pp. 494–499. Archaeopress, Oxford (2004)
6. Beex, W., Peterson, J.: The Arminghall henge in space and time: how virtual reality contributes
to research on its orientation. In: der Stadt Wien, M. (ed.) Enter the Past – The E-Way into
the Four Dimensions of Cultural Heritage CAA 2003 Proceedings. BAR International Series,
vol. 1127, pp. 490–493. Archaeopress, Oxford (2004)
7. Wendrich, W.Z., Bos J.E.M.F., Pansire, K.M.: VR modeling in research, instruction,
presentation and cultural heritage management: the case of Karanis (Egypt). In: Ioannides,
M., Arnold, D., Niccolucci, F., Mania, K. (eds.) The Evolution of Information Communication
Technology in Cultural Heritage. Where Hi-Tech Touches the Past: Risks and Challenges for
the 21st Century, Nicosia, pp. 225–230 (2006)
8. Meyer, Ė., Grussenmeyer, P., Perrin, J., Durand, A., Drap, P.: Intra-site level cultural heritage
documentation: combination of survey, modeling and imaginary data in web information
systems. In: Ioannides, M., Arnold, D., Niccolucci, F., Mania, K. (eds.) The Evolution of
Information Communication Technology in Cultural Heritage. Where Hi-Tech Touches the
Past: Risks and Challenges for the 21st Century, Nicosia, pp. 129–134 (2006)
9. Mourkoussis, N., Mania, K., Petridis, P., White, M., Rivera, F., Pletinckx, D.: An analysis of
the effect of technological fidelity on perceptual fidelity. In: Proceedings of the IEA 2006
(International Ergonomics Association), 16th World Congress on Ergonomics, The Hague,
Netherlands (2006)
10. Petridis, P., Pletinckx, D., Mania, K., White, M.: The EPOCH multimodal interface for
interacting with digital heritage artefacts. In: Zha, H., Pan, Z., Thwaites, H., Addison, Alonzo,
C., Forte, M. (eds.) VSMM 2006. LNCS, vol. 4270, pp. 408–417. Springer, Heidelberg (2006).
doi:10.1007/11890881_45
11. Pletinckx, D., Silberman, N.A., Callebaut, D.: Heritage presentation through interactive
storytelling: a new multimedia database approach. J. Vis. Comput. Anim. 14(4), 225–231
(2003). Wiley
12. Irrlicht Engine, http://irrlicht.sourceforge.net/
13. irrKlang Audio Library, http://www.ambiera.com/irrklang/
14. Newton Game Dynamics Physics Engine, https://code.google.com/archive/p/newton-
dynamics/downloads
15. LibCurl Multiprotocol File Transfer Library, https://curl.haxx.se/libcurl/
16. Ffmpeg Audio/Video conversion tool, https://ffmpeg.org/
17. Autodesk Maya 2012, http://www.autodesk.com/products/maya/overview
18. Adobe Photoshop CS 3 Extended, http://adobe-photoshop-cs3-extended.soft112.com/
19. Azuma, R.: A survey of augmented reality. Presence: Teleoperators Virtual Environ. 6(4),
355–385 (1997)
20. Luebke, D., Reddy, M., Cohen D.J., Varshney, A., Watson, B., Huebner, R.: Level of Detail
for 3D Graphics. In: Barsky, B.A. (ed.) University of California, Berkeley. Morgan Kaufmann
Publishers, Elsevier Science (USA) (2003)
DICE: Digital Immersive
Cultural Environment
Stelios C.A. Thomopoulos(&), Adam Doulgerakis, Maria Bessa,

Konstantinos Dimitros, Giorgos Farazis, Eftichia Georgiou,
Tassos Kanellos, Christina Karafylli, Maria Karafylli,
Dimitris M. Kyriazanos, Vassilios I. Kountouriotis,
Vassilis Lampropoulos, Christos Margonis, Christos Maroglou,
Dionisis Motos, Alexandra Papagianni, Manolis Paterakis,
Katerina Skroumpelou, Giorgos Konstandinos Thanos,
Ino-Eleni Theodorou, Christina Phobe Thomopoulos,
Panagiotis Tsimpiridis, Dimitris Zacharakis, and Andreas Zalonis
Integrated Systems Laboratory, Institute of Informatics and Telecommunications,

National Center for Scientific Research “Demokritos”, 15310 Athens, Greece
{scat,adoulgerakis,mbessa,k.dimitros,gfarazis,
efgeorgiou,tassos.knl,ckarafilli,mkarafilli,dkyri,
bkoutoutiotis,vlamp,chris.margonis,cmaro,d.motos,
alexpapagianni,mpaterakis,ksroumbelou,giorgos.thanos,
itheodorou,cpt,ptsimpiridis,dzacharakis,
azalonis}@iit.demokritos.gr
Abstract. A Digital Immersive Cultural Environment (DICE) consists of a

VR/AR (virtual & augmented reality) platform, a CMS (Content Management
System) and a GIS (Geographic Information System) for geo-referencing both
space and content and for providing immersive navigation in the VR/AR space.
Such a platform offers 3D reconstruction of space, geo-coding of the virtual
space in actual geographic coordinates and overlay capabilities with real geo-
referenced space. The coupling of a CMS with a GIS associates information and
data with the geographic coordinates of the VR/AR space, thus making available
content on demand in accordance to spatial point of interest within the VR/AR
space and immersive navigation in both VR and AR spaces. Incorporation of
simulation platforms, localization technologies, motion tracking technologies
and VR UI’s, creates a fully interactive immersive VR/AR ecosystem, beyond
the state of the art, with augmented capabilities in education, training, enter-
tainment, content creation, etc.
Keywords: Virtual reality Augmented reality Digital cultural space GIS

Indoor localization Immersive navigation Simulation
From the second author, names are in alphabetical order.

DOI: 10.1007/978-3-319-48496-9_61
DICE: Digital Immersive Cultural Environment 759
1 Introduction
The concept of the DICE as a means of creating a sustainable ecosystem for the
development of 3D digital (virtual and augmented reality) immersive cultural envi-
ronments for a wide range of activities, from education, to research, to entertainment
and leisure, was first introduced by S.C.A. Thomopoulos in 2014 [1] and was refined in
2015 [2]. The DICE concept was motivated and initially introduced by the need for the
creation of a comprehensive and sustainable ecosystem for smart specialization in the
area of Culture and Tourism. As it can be seen in Fig. 1, the “Wheel on innovation” (to
be referred to as WOI in what follows) consists of three sectors (or elements) that it is
necessary to work in concertation in order to maintain its spin.
Fig. 1. The “wheel of innovation” for the creation of a DICE ecosystem for Culture & Tourism
The most critical element of WOI is the Ecosystem of Stakeholders & Beneficiaries
that includes (and the list is not exhaustive): Creators of Digital Immersive Environ-
ments (e.g. architects, digital artists, digital designers, etc.), Historians, Archaeologists,
Humanity & Social Scientists, Behavioral Scientists, Information & Communication
Technologies, Physical Scientists, Mathematicians & Analysts, Archaeological Sites &
Museums, Touristic Destinations, Regional Governments, Network Operators & ISPs,
760 S.C.A. Thomopoulos et al.
etc. All these together acting as the scientific, technological and business catalysts will
enable the creation of a DICE platform capable of delivering the functionalities that are
necessary for an immersive cultural environment.
The second critical element of WOI is the Creative Innovation Environment: In
order to coherently and over time coordinate the development in the complex WOI
ecosystem, it is necessary to have in place a Center of Excellence in Creative
Immersive Design, Arts & Games, Human Behavior Modeling, Computational Soci-
ology and Simulation, and Creative Story Boarding. Furthermore, an SME Incubator is
necessary as a complementary ingredient of the Center of Excellence in order to
facilitate, expedite and catalyze the fast transition of the technological developments
into the market and the expedient extraction of the added value within them.
Last, but not least, the third pillar “Shaping of Disruptive Technologies” is required
to foster an environment for creating disruptive technologies with exponential benefits
to the economy and the generation of new added value through a Holistic Immersive
Tourism approach, Serious Interactive 3D Educational Games, and a Global Partici-
patory Creativity Platform enabling technological crowd sourcing, and, as an ultimate
goal, the Design of the Living Space of the Future. Due to the limitation in space and
the focus on the educational and collaborative aspects of DICE, we will only present
our efforts in developing the fundamental components that can make DICE a reality
and how these can lead to measurable educational and collaborative results. The
financial, economic and marketing aspects of DICE are not covered in this paper.
2 Building the DICE Ecosystem: Components and Building

Blocks
Building the DICE Ecosystem requires a number of technological elements ad com-

ponents in place in order to ensure the capabilities that are needed for:
– Generation of 3D VR/AR spaces
– Content creation and content management
– VR/AR space and related content geo-referencing
– Storyboarding and immersive navigation
– Automated Content Adaptation
– Content DB specification
– Narrative and Content Delivery Method Authoring Tools
– Simulation, gamification and visualization tools
– UI/UX for intuitive and immersive VR/AR interfaces and devices
– Crowdsourced Content Aggregation, Validation and Recommendation
– Collaborative research, education and co-creation tools
– Personalized access to content based on preferences, user profile, disabilities, etc.
– Personalized Storytelling
– Social Media Platform Integration
– Common Specifications towards a unified UI providing access to all content types,
including specifications concerning accessibility limitations due to disabilities.
The above are some of the elements that are required in order to be able to accom-
modate the stakeholders of the DICE ecosystem as delineated in Fig. 1. Analyzing the
above, possibly non-exhaustive list of requirements for a DICE ecosystem as a tech-
nological, collaborative, inclusive and immersive platform, the following key building
blocks emerge:
#1 A platform for the creation of 3D virtual environments that it is either based on
commercial, popular 3D design platforms or interfaces with them, a set of tools
and plague-ins for embedding content (in the form of text, audio, video,
animations, immersive navigation, etc.), a gamification engine, either as an
internal engine or a plague-in to a commercial engine, an interface with a
powerful simulation engine for the creation of necessary animations within the
immersive VR/AR environment, and powerful UI/UX’s to guarantee integration
of VR equipment and devices for immersive experiences.
#2 A platform for geo-referencing 2D/3D architectural and geolocating indoor and
outdoor 2D/3D and their content so that they can be georeferenced on a map, thus
allowing information retrieval based on location information, access and
navigation in VR and AR spaces using locative services, and geographical
alignment of virtual 2D/3D reconstructions with actual 2D/3D buildings and
artifacts.
#3 A platform for allowing general purpose simulations, including human behavior,
motion and dynamics of human, animal and physical system, and device
modelling and simulation to allow scenario-based animations, immersive
interactions of the users within the platform through avatars, gamification and
interactivity.
#4 A platform of a distributed simulation environment for coupling different
simulators of events, spaces, behaviors, and physical phenomena or man-made
mechanisms, devices or processes, and embedding the output of these simulators
in the DICE environment.
#5 A platform for crowd-sourcing (multimedia) content, allowing aggregation,
validation and recommendation of the crowd-sourced content and interface for
integration with different social media platforms.
#6 A platform for collaborative research, education and training to include
co-creation tools, personalized access to content based on preferences, user
profile, disabilities, etc., personalized storytelling, and UI/UX for intuitive and
immersive VR/AR interfaces and devices.
ISL has already developed the first four platforms required for creating the DICE
ecosystem, and is in the process of completing platforms #5 and #6. The first four
platforms are, namely:
(a) i-Guide, a 3D VR/AR creation and navigation platform for building block #1 [3];
(b) wayGoo, a platform for geolocating and managing indoor and outdoor 3D spaces
and content within for building block #2 [4];
(c) OCULUS Sim distributed simulation platform consisting of two key components:
a. i-Crowd a general purpose simulation engine for behavior, motion and
dynamics of human, animal and physical system, and device modelling and
simulation for building block #3 [5]; and
b. Pyrones, a distributed simulation environment of coupling simulators of dif-

ferent phenomena and events with different visualization capabilities as
required by different application domains for building block #4 [6]; and
(d) i-Rescue 3D immersive & interactive gamification platform for building block #4
[6].
have reached a stage of maturity between TRL7 and TRL9 (Technological Readiness
Level 7 to 9) and are described next, whereas the other two for #5 and #6 are in the
process of development.
In addition to its contributions through the each one of the six different platforms
that constitute its key building blocks, DICE offers an integrated platform that goes
beyond the current state of art in digital immersive environments for culture, education,
entertainment, collaborative working, and tourism.
The first four of the six platforms are described briefly in the next Section whereas
examples of their usage in the context of promoting the DICE ecosystem are given in
Sect. 5.
3 Description of the Four DICE Platforms
The four platforms that are described next have been built in a modular way so that can
either be used as separate applications or in conjunction with each other as building
blocks of a modular and distributed DICE architecture that integrates various plat-
form components to deliver an ecosystem that accommodates the needs and require-
ments of all stakeholders involved in the ecosystem.
(a) i-Guide [3], a 3D VR/AR creation and navigation platform: i-Guide aims at
enhancing the user experience as a “digital” visitor to archaeological sites and mon-
uments. Through the detailed 3D reconstruction of the monument the visitor is offered
a first person navigation in the original form of the monument. The platform offers a
number of features that allow the systematic development of immersive digital cultural
environments and digital tours that enable the creator of such spaces to design them in a
non-ad hoc but methodical manner, thus accelerating their productions and allowing for
a signature UI-UX look-and-feel. Knossos i-Guide is the first 3D immersive “digital”
visit to the Palace of Knossos in Crete entirely developed within the i-Guide platform.
Figure 2 depicts the modular elements of the Knossos i-Guide platform that allow the
management of any multimedia digital content in a uniform way by simply replacing
the core 3D VR Knossos reconstruction by any other, actual or imaginary, 3D VR
reconstruction and a simple content adaptation of the content to the relevant artifact.
(b) wayGoo [4], a platform for geolocating and managing indoor and outdoor 3D
spaces and content within: wayGoo is a platform for Geolocating and Managing indoor
and outdoor spaces and content with multidimensional indoor and outdoor Navigation
and Guidance. Its main components are a Geographic Information System, a back-end
server, front-end applications and a web-based Content Management System (CMS). It
constitutes a fully integrated 2D/3D space and content management system that creates
a repository that consists of a database, content components and administrative data
Fig. 2. Knossos i-Guide platform
(Figs. 3 and 4). wayGoo can connect to any third party database and event management
data-source. The platform is secure as the data is only available through a Restful web
service using https security protocol in conjunction with an API key used for
authentication. To enhance user experience, wayGoo makes the content available by
extracting components out of the repository and constructing targeted applications. The
wayGoo platform supports geo-referencing of indoor and outdoor information and use
of metadata. It also allows the use of existing information such as maps and databases.
The platform enables planning through integration of content that is connected either
spatially, temporally or contextually, and provides immediate access to all spatial data
through interfaces and interactive 2D and 3D representations. wayGoo constitutes a
mean to document and preserve assets through computerized techniques and provides a
system that enhances the protection of your space, people and guests when combined
with wayGoo notification and alert system. It constitutes a strong marketing tool
providing staff and visitors with an immersive tool for navigation in indoor spaces and
allowing users to organize their agenda and to discover events through wayGoo event
scheduler and recommendation system. Furthermore, the wayGoo platform can be used
in Security applications and event management, e.g. CBRNE incidents, man-made and
natural disasters, etc., to document and geolocate information and sensor data (off line
and real time) on one end, and offer navigation capabilities in indoor and outdoor
spaces. Furthermore, the wayGoo platform can be used for the creation of immersive
environments and experiences in conjunction with VR/AR (Virtual & Augmented
Reality) technologies.
Fig. 3. wayGoo platform server-side working environment
(c) i-Crowd [5], a general purpose simulation engine: Initially designed as a crowd
behavior simulator for simulating, testing and visualizing evacuation scenarios, the
i-Crowd Simulation platform has evolved into a complete domain-independent
agent-based behavior simulator with an emphasis on crowd behavior and building
evacuation simulation. Under continuous development, it reflects an effort to imple-
ment a modern, multithreaded, data-oriented simulation engine employing latest
state-of-the-art programming technologies and paradigms. It is based on an extensible
architecture that separates core services from the individual layers of agent behavior,
offering a concrete simulation kernel designed for high-performance and stability. Its
primary goal is to deliver an abstract platform to facilitate implementation of several
Agent-Based Simulation solutions with applicability in several domains of knowledge,
such as: (i) crowd behavior simulation during [in/out] door evacuation; (ii) non-player
character AI for game-oriented applications and gamification activities; (iii) vessel
traffic modeling and simulation for Maritime Security and Surveillance applications;
Fig. 4. UI/UX mobile implementation of wayGoo for different application domains
(iv) urban and highway traffic and transportation simulations; (v) social behavior
modeling and simulation; and in general, (vi) behavior, motion and dynamics of
human, animal and physical system and device modelling and simulation platform
(Fig. 5).
Some notes on the simulator:
• The simulator itself is designed as an abstract Agent-Based System, and it is based
on the Entity-Component design paradigm.
• Τhe Entity is the main architectural unit of the simulator, being the only object that
resides directly inside the simulation kernel.
• Complex behaviors are built by populating Entities with Components provided by
Layers that reside outside the core, as plug in systems.
• Thus, the main processing kernel that deals with resource allocation and processing
synchronization is separated from the individual behavior implementations that
usually deal with higher level functionalities (steering, path finding, intelligence,
communications etc.).
Fig. 5. i-Crowd simulator platform
(d) PYRONES [6], developed in the context of modeling and simulating large
structural fires continue that pose a great threat towards human life and property,
PYRONES has become a modular, distributed and highly integrated platform for
building and human behavior modeling platform under any catastrophic events. Due to
the complexity and non-deterministic characteristics of a building fire disaster, it is not
a straightforward task to assess the effectiveness of fire protection measures embedded
in the building design, planned evacuation strategies and potential modes of response
for mitigating the fire’s consequences. Additionally, there is a lack of means that
realistically and accurately recreate the conditions of building fire disasters for the
purpose of training personnel in order to be sufficiently prepared when vis-a-vis with
such an environment.
The PYRONES system has been developed to address all these aspects through a
comprehensive approach that relies on accurate and realistic computer simulations of
the individual phenomena and their interactions. PYRONES offers innovative tools and
services to strategically targeted niches in two market domains. In the domain of
building design and engineering, PYRONES is seamlessly integrated within existing
engineering Building Information Modelling (BIM) workflows and serves as a building
performance assessment platform, able to evaluate fire protection systems. On another

front, PYRONES penetrates the building security management market, serving as a
holistic training platform for specialists in evacuation strategy planning, firefighters and
first responders, both at a Command & Control and at an individual trainee level. The
overall PYRONES architecture, simulation and visualization modes are shown in
Figs. 6 and 7.
Fig. 6. PYRONES distributed modular architecture
iRescue is an immersive 3D game platform developed as an active game (ex-

ergame), using the Kinect sensor for the tracking of the player’s movement and
position. It also available for passive game playing for various game and mobile
platforms.
Figure 8 shows a scene for an i-Rescue game that takes place at an airport after a
catastrophic event has occurred. iRescue immersive 3D game places the player within
an airport after a disastrous event. As a member of an elite paramedic corps, the
player’s objective is to guide the evacuating panicked crowd towards a safe pathway.
Diverse and unpredicted simulated crowd behaviours, gradually advancing crowd
density and difficulty, as well as special actions, aim to capture the player’s interest in
an entertaining and challenging experience. Due to control system malfunctioning
during landing, the n crashes on the main terminal building causing serious damages to
the airport building structure. Rifts and debris are everywhere raising havoc among the
passengers.
Only an elite paramedic can deal with the situation and is the first to respond to the
site by motorcycle, in order to guide the passengers to safely evacuate the building
without falling in a rift. The main objective of the player, is to save as many persons as
Fig. 7. PYRONES simulation and visualization modes

Fig. 8. A scene for the i-Rescue airport rescue game
she can, in order to achieve a better score and rank. Using the Kinect sensor for the
tracking of the player’s movement and position, the player interacts with the VR
environment in a natural way trying to save as many people as possible.
For additional information about the different platforms DICE is made up of, the
reader is referred to [15] as well.
4 DICE Architecture: The DICE System/Platform

and Innovation
The DICE architecture aims at the development of an innovative, integrated yet

modular platform for improving access to multimedia content and facilitating the
creation of 3D digital immersive environments, by using the latest advances in visual
representation technologies and a combination of advanced and novel features that
facilitate deeper cultural learning and understanding, by immersing visitors into social
aspects of the culture associated with the exhibition.
The digital content (both tangible and intangible) is: (a) organized using various
multi-dimensional hyper structures and topologies, such as neural networks, hyper-
spaces and clusters overlaid on the 4D space-time continuum; (b) related through
dynamic semantic synapses; (c) presented/curated through a variety of curating, sto-
rytelling and editing tools for organizing and curating a collection of digital assets with
a specific, yet dynamically experienced and interactive context; (d) visualized through a
variety of information visualization techniques, VR/AR technologies and affective
interfaces; and (e) experienced through a variety of modalities that allow remote and
on-site access, first and third person visits, individual and group tours, and direct
interactivity through social media or indirect through gamification and avatars.
The content is delivered through an immersive mixed reality user experience,

referring both to the relevant physical and virtual cultural spaces and artefacts. The
cultural assets are geo-temporally linked to their context and accessed through virtual
(for remote access) and/or augmented reality (for in-situ access) technologies. The
platform enables access to reconstructions of shattered or broken artefacts, identifica-
tion and reunification of contextually related assets that have been scattered across
multiple museum collections.
DICE exploits the technology, not only towards the optimum presentation of the
digital content, but also towards societal and educational dimension. As presented in
various studies, VR holds considerable potential as a powerful medium for learning: it
gives users the ability to carry out tasks that could be difficult or impossible in the “real
world” due to constraints such as cost or location; allows for continuing and growing
social interactions which can serve as a basis for collaborative education; allows indi-
viduals – especially children – to learn from non-symbolic first-person experience.
Along these lines, the platform combines VR/AR with a novel gamification layer that
allows the VM visitor to participate in stories and events related to the VM collection
exhibits: the visitor may select a virtual character and communicate with other visitors
(virtual or real); interact with characters of that era and immerse into their everyday life;
learn a skill and engage in the creation of an artefact or product of that era; play games
based on historical facts or on fictional histories related to the exhibits. This gamification
layer is built with the engagement of cultural heritage professionals and learning experts
in order to effectively achieve the education of all VM visitors on the subject matter.
Social Media and blogging tools (such as Facebook and Twitter) are integrated into
the platform to invite end-users to share their personal experience, effectively con-
verting it to a virtual meeting and learning place. The DICE is a space that can offer
visitors – and especially young people – opportunities for participation, sharing, con-
versation, interaction that allows a more effective contextualization and interpretation
of cultural assets and assist professionals in assessing the understanding of cultural
content in VR/AR environments by broader audiences.
Within this framework, the main attributes of the DICE architecture/platform, as
shown in Fig. 9, are:
(1) State-of-the-art front-end components which incorporate the best available tech-
nology that will assure the optimum visual quality and user experience, using all
relevant ICT tools (VR - Oculus Rift/Google Cardboard, AR – Vuforia/Wikitude,
gesture-based Interactivity, 3D modelling, high quality lighting and shading,
Spatial 3D sound, animations);
(2) An innovative and intuitive authoring tool that provides unique ways to organize,
relate, present/curate, visualize and experience the digital content and information,
allowing the efficient usage and/or further development of the platform from the
cultural heritage professionals.
(3) A flexible and scalable back-end component that can synthesize diverse information
from various sources, by also incorporating semantic tools that allow the use of
diverse information from different databases, and also assists the front-end com-
ponents in the creation of visual narratives, histories, games, matched to the cura-
tor’s objectives, the individual visitor’s characteristics, and the available content;
Fig. 9. DICE Architecture adapted to the creation and curation of a Virtual Museum
(4) A gamification layer for learning and education purposes – the developed games
will be incorporated into the platform as additional assets for exploitation by the
curator.
(5) The holistic approach on personalized narration/storytelling and behavior-based
adaptation with the inclusion of a large number of available information for the
individual visitor, including social media interactions, sentiment analysis, per-
sonal information, experience from prior visits;
(6) Social Media integration.
5 DICE Use Cases in Education and Collaborative Working
Next we present three use cases of pilot projects that took place within the DICE
ecosystem.
Use-Case 1: Knossos i-Guide Education and Gaming Platform [3]
Knossos i-Guide (Fig. 2), has been developed as a platform for education and
gaming aiming at achieving an educational immersive tour guide in the virtually
reconstructed archaeological palace of Knossos enriched with animated content, nar-
rations, manual and automatic tour guides, validated rich historical and archaeological
content, info-points within the tour, overlay of pictures of today site through the
patent-pending “gelatin™ mechanism,” educational digital art games and puzzles, and
related adventure games. The Knossos i-Guide platform has been tested in various
(informal) educational events and activities such as the:
(a) Athens Science Festival (ASF 2015/2016 http://www.athens-science-festival.gr/
en/) [7];
(b) Researcher’s Night (2013, 2014, & 2015), an EU instituted annual event
throughout all member-states to celebrate research achievements and open
research organizations and their research work to the public (https://www.iit.
demokritos.gr/news/ren2015_post) [8];
(c) “Archaeological Workout” workshop organized in the context of the annual
event “Archaeological Dialogues” in 2015 & 2016, whereby digital and con-
ventional approaches to learning about history and archaeology were coupled
harmoniously to achieve better understanding and appreciation of culture, art and
history (http://archaeological-workout-workshop.iit.demokritos.gr/) [9]; and
(d) SMART Camp on digital art, a fast-track, 5-day summer camp organized and
taught by researchers from ISL to students 13-18 years old, where students are
taught how to use design tools for creating virtual spaces and animation, and how
to present their digital creations using projection mapping techniques (http://
smart-camp.iit.demokritos.gr/) [10].
Use-Case 2: PYRONES Collaborative Distributed Simulation Platform for
Engineering Design and Security Training [6, 11]
PYRONES is a collaborative distributed real-time simulation platform that is
motivated by the threat that structural fires continue to pose towards human life and
property and the need to understand how these fires interact with the materials building
are built of and how they propagate; what health threat these fires impose on tenants of
a building; how effective escape mechanisms and design are; how evacuation plan
function; and last, but not least, how active fire extinguishing mechanisms, e.g.
sprinklers, retardants, etc., help in preventing the fire from reaching life threatening
levels. Due to the complexity and non-deterministic characteristics of a building fire
disaster, it is not a straightforward task to assess the effectiveness of fire protection
measures embedded in the building design, planned evacuation strategies and potential
modes of response for mitigating the fire’s consequences. The PYRONES platform
(Figs. 6 and 7) has been developed to address all these aspects through a compre-
hensive approach that relies on accurate and realistic computer simulations of the
individual phenomena and their interactions. The platform offers collaboration at dif-
ferent levels, as presented in Fig. 10: (a) at the simulation level by offering integration
of simulation modules from different providers and for different eventualities; (b) at the
integration level by offering plug-ins for commercial software tools, such as Revit from
Autodesk; and (c) at the application/market level: to engineering for building designing
and testing and to security for training of first responders.
The PYRONES simulation platform was tested by performing a simulation of the
Station Nightclub structural fire incident, which took place in West Warwick, Rhode
Island in 2003. The specific event is considered to be a benchmark simulation scenario
as it was one of the most lethal and well-studied fire accidents, which caused 96 deaths.
The National Institute of Standards and Technology (NIST), in the report of the
technical investigation of the incident, estimates the maximum safe population limit for
the specific building to be 420. Based on press reports for the number of occupants at
the station nightclub the night of the incident, the testing of the PYRONES system
considered a distribution of 460 people within the building.
Fig. 10. PYRONES Overall architecture and collaborative environment
Three scenarios were tested [11]. While in the first scenario implementation fire
advancement and crowd evacuation were examined as two independent phenomena, in
the other two scenarios these phenomena were considered as interconnected processes.
The first scenario implementation followed the scenario of the first simulation con-
ducted by NIST, which did not couple fire and evacuation and the results of the
PYRONES simulation were compared to the NIST’s simulation results. The second
scenario coupled fire and evacuation simulation and the results of the PYRONES
simulation were compared to the actual incident’s results as reported by Rhode Island
Attorney General’s Office. The third scenario coupled fire with evacuation and it was
further assumed that sprinklers were installed in the building. This scenario was used to
assess the effectiveness of a plausible enhancement in the building fire security
measures.
The results of the PYRONES simulations were compared to the NIST’s simulation
results, in terms of temperature, carbon monoxide and heat flux.
• The results of the execution of the scenario with a delay of 40 s under predict the
actual number of victims by a mere 9 % (the number of predicted fatalities was
87 ± 3), providing a low error margin.
• The estimated number of people remaining at 90 s for the second scenario (with
fire) was 295 which is considerably more than that in scenario 1 (228 people,
without fire). Thus, it is important to couple fire and crowd, injuries models in order
to calculate the average safe evacuation time.
• The main threats to the occupants were the toxic gases and the heat release. Most of
the locations in the simulated station nightclub became untenable due to the toxic
gasses concentrations14 (CO concentration exceeded 3200 ppm after 80 s), as well
as the temperature9 (temperatures exceeded the 120 °C threshold after 60 s). This
conclusion is in accordance with the conclusions drawn in the NIST report.
• Most of the fatalities in PYRONES simulation were caused by the toxic gases
(specifically most of the fatalities were caused by the concentration of CO,
approximately 62 % of fatalities). This also agrees with the NIST report, where
most fatalities are described to be caused due to toxic gas inhalation.
Use-case 3: wayGoo Collaborative Indoor Space Creation and Content Geo-

referencing Platform [4]
wayGoo is a platform for Geolocating and Managing indoor and outdoor spaces and
content with multidimensional indoor and outdoor Navigation and Guidance. Its main
components are a Geographic Information System, a back-end server, front-end appli-
cations and a web-based Content Management System (CMS). It constitutes a fully
integrated 2D/3D space and content management system that creates a repository that
consists of a database, content components and administrative data. wayGoo can connect
to any third party database and event management data-source. The platform is secure as
the data is only available through a Restful web service using https security protocol in
conjunction with an API key used for authentication. To enhance user’s experience,
wayGoo makes the content available by extracting components out of the repository and
constructing targeted applications. The wayGoo platform supports geo-referencing of
indoor and outdoor information and use of metadata. It also allows the use of existing
information such as maps and databases. The platform enables planning through inte-
gration of content that is connected either spatially, temporally or contextually, and
provides immediate access to all spatial data through interfaces and interactive 2D and
3D representations. wayGoo constitutes a means for documenting and preserving assets
through computerized techniques and provides a system that enhances the protection of
your space, people and guests when combined with wayGoo notification and alert sys-
tem. It constitutes a strong marketing tool providing staff and visitors with an immersive
tool for navigation in indoor spaces and allowing users to organize their agenda and to
discover events through wayGoo event scheduler and recommendation system. Fur-
thermore, the wayGoo platform can be used in Security applications and event man-
agement, e.g. CBRNE incidents, man-made and natural disasters, etc., to document and
geolocate information and sensor data (off line and real time) on one end, and offer
navigation capabilities in indoor and outdoor spaces. Furthermore, the wayGoo platform
can be used for the creation of immersive environments and experiences in conjunction
with VR/AR (Virtual & Augmented Reality) technologies.
The wayGoo platform has been used to geolocate various events, exhibits and fairs,
such as the Mobile World Congress 2015 in Barcelona, Spain; the campus of NCSR
Demokritos, the 2nd Hellenic Forum for Science and Innovation, the Researcher’s
Night exhibit in 2014 and 2015, and the Athens Science Festival 2015 & 2016.
A virtual tour of these event can be downloaded from both AppStore and PlayStore by
searching for wayGoo and downloading the mobile apps.
The use of wayGoo in the context of DICE is quite apparent for it allows the
accurate georeferenced of content and building (actual and virtual), thus enabling:
(a) immersive navigation in a cultural space and location-based services associated with
a virtual or physical visit or tour of the site; (b) innate geographical alignment between
virtual reality and actual/physical reality artifacts for the creation of truly immersive
VR/AR experiences.
Summing up, the three use cases described above offer precious experience for
addressing similar issues, such as immersive and interactive education in culture and
the arts, impact analysis of hazardous events in cultural monuments and means to
address and mitigate them, and content management and navigation in cultural spaces
in conjunction with the use of VR/AR technologies.
Fig. 11. DICE Ecosystem block diagram implementation with the key components shown as
modular building blocks
6 Conclusions
The DICE ecosystem described in Fig. 1 is gradually being translated into a techno-
logical platform (of platforms) through a modular approach, by developing a dis-
tributed architecture that allows major components of the DICE puzzle, each built
separately and in a modular way, to be integrated into one common DICE Platform of
Platforms (PoP). ISL has developed already four critical platforms described in this
paper and tested them in a variety of pilot and commercial use cases. Furthermore,
using its prior Systems of Systems (SoS) design and implementation experience from
the PERSEUS demonstration project [12, 13], ISL is integrating all these platforms into
one common but flexible DICE platform while working to develop and complete the
remaining building blocks for content creation, storyboarding and storytelling, gami-
fication, personalization and crowd sourcing integration (Fig. 11).
Acknowledgments. a. The work and results presented in this paper has been supported by the
following research projects [14]:
AF3: Advanced Forest Fire Fighting, Grant Agreement No. 607276, funded under
FP7-SECURITY - Specific Programme “Cooperation”: Security of the European Commission.
FLYSEC: Optimising time-to-FLY and enhancing airport SECurity, Grant Agreement
No. 653879, funded under programme H2020-EU.3.7. – “Secure societies - Protecting freedom
and security of Europe and its citizens” of the European Commission.
PYRONES: PYRo-mOdelliNg and Evacuation Simulation system: This project was sup-
ported by the Programme: NSRF/ESPA (EPAN II), Contract Number: ISR 2866, funded by the
General Secretariat for Research and Technology (GGET), Hellenic Ministry of Education,
Religious Affairs, Culture and Sports, and the Israeli Industry Center for R&D, Matimop, under
the “Greece-Israel Bilateral R&T Cooperation 2013–2015”.
SYNAISTHISI: Intelligent data collection and processing platform for energy efficient
applications, Contract No. 2012 KRHPIS, Hellenic General Secretariat of Research & Tech-
nology, Duration: 32 months (02/04/2013 – 12/31/2015), Co- financed by Greece and the
European Union- European Regional Development Fund (ERDF), under the O.P. “Competi-
tiveness and Entrepreneurship” (OPCE II) and of R.O.P Attica, R.O.P Macedonia- Thrace.
GRECORISK: Hellenic Natural-Hazards Risk-Mitigation System of Systems, funded under
Programme: SYNERGASIA 2011, Contract Number: 11SUN 8 1696, 2013–2015.
TASS: Total Airport Security System (FP7-SEC-2010- Contract No. 241905) research project
funded by the European Commission and, in part, by the General Secretariat of Research and
Technology (GSRT) of the Ministry of Education and Religious Affairs, Culture and Sports,
Greece.
PERSEUS: Protection of European seas and borders through the intelligent use of surveil-
lance, FP7-SEC-2010.3.1-1-Contract No. 261748.
b. The co-authors are all colleagues at the Integrated Systems Laboratory (ISL) and have
contributed towards the enhancement of the DICE concept and the implementation of the DICE
platform and its various components as follows:
• DICE Platform Architecture: Tassos Kanellos, Adam Doulgerakis, Maria Bessa, Eftichia
Georgiou and Andreas Zalonis.
• i-Guide 3D VR/AR visualization, tour guide, navigation platform & story boarding plat-
form: Giorgos Farazis, Christos Maroglou, Panagiotis Tsimpiridis, Alexandra Papagianni,
Ino-Eleni Theodorou and Christina Phobe Thomopoulos.
• wayGoo indoor geocoding & navigation platform: Christina Karafylli, Maria Karafylli,
Dionisis Motos, Vassilis Lampropoulos, Manolis Paterakis, Christos Maroglou, Konstantinos
Dimitros, Dimitris Zacharakis, Giorgos Konstandinos Thanos, Katerina Skroumpelou and
Dimitris M. Kyriazanos.
• i-Crowd human behavior ssimulator: Vassilios I. Kountouriotis, Manolis Paterakis
• PYRONES distributed simulation platform: Vassilios I. Kountouriotis, Manolis Paterakis,
Tassos Kanellos and Adam Doulgerakis.
• i-Rescue gamification platform: Adam Doulgerakis, Tassos Kanellos, Betty Evangelinou and
Dafni Polyzos.
References
1. https://prezi.com/mvja6fyp3ldq/copy-of-2014-copyright-isl-ncsr-demokritos-all-rights-
reserv/
2. https://prezi.com/embed/zx-bzzmmvy5s/
3. Knossos i-Guide video teaser v4 and educational games. https://vimeo.com/105324264,
https://vimeo.com/120785074, https://vimeo.com/120784644
4. Thomopoulos, S.C.A., Karafylli, C., Karafylli, M., Motos, D., Lampropoulos, V., Dimitros,
K., Margonis, C.: WayGoo: a platform for geolocating and managing indoor and outdoor
spaces. In: Proceedings of SPIE, Baltimore, USA, 18–20 April 2016, vol. 9842 (2016)
5. Kountouriotis, V.I., Paterakis, M., Thomopoulos, S.C.A.: iCrowd: agent-based behavior
modeling and crowd simulator. In: Proceedings of SPIE, Baltimore, USA, 18–20 April 2016,
vol. 9842 (2016)
6. Kanellos, T., Doulgerakis, A., Georgiou, E., Kountouriotis, V.I., Paterakis, M., Tho-
mopoulos, S.C.A., Pappou, T., Vrahliotis, S.I., Rekouniotis, T., Protopsaltis, B., Rozenberg,
O., Livneh, O.: PYRONES: pyro-modelling and evacuation simulation system. In:
Proceedings of SPIE, Baltimore, USA, 18–20 April 2016, vol. 9842 (2016)
7. Athens Science Festival 2015. https://vimeo.com/126392686
8. Researcher’s Night (2013, 2014 and 2015). https://www.iit.demokritos.gr/news/ren2015_
post
9. Archaeological Workout. https://vimeo.com/159943241
10. SMART Camp on digital art. https://vimeo.com/150669087, https://vimeo.com/161189858
11. PYRONES use case scenario simulation video. https://vimeo.com/137834302, https://vimeo.
com/163534330, https://vimeo.com/154047986
12. PERSEUS: Protection of European seas and borders through the intelligent use of
surveillance. http://www.perseus-fp7.eu/, https://vimeo.com/135040978, https://vimeo.com/
118606731
13. Kanellopoulos, S., Katsoulis, S., Motos, D., Lampropoulos, V., Margonis, C., Dimitros, K.,
Thomopoulos, S.C.A.: OCULUS Sea: integrated maritime surveillance platform. In: SPIE
Defense and Security, Signal Processing, Sensor/Information Fusion, and Target Recogni-
tion XXIV, Paper 9474-23, 21 April 2015
14. https://www.iit.demokritos.gr/isl/projects
15. http://claret.gr/
Project Paper: The New Era of
Museums and Exhibitions: Digital
Engagement and Dissemination
Imaging Novecento. A Mobile App
for Automatic Recognition of Artworks
and Transfer of Artistic Styles
Federico Becattini, Andrea Ferracani(B) , Lea Landucci, Daniele Pezzatini,

Tiberio Uricchio, and Alberto Del Bimbo
NEMECH - New Media for Cultural Heritage, University of Florence,

Viale Morgagni 65, 50124 Firenze, Italy
{federico.becattini,andrea.ferracani,lea.landucci,daniele.pezzatini,
tiberio.uricchio,alberto.delbimbo}@unifi.it
http://nemech.unifi.it
Abstract. Imaging Novecento is a native mobile application that can be

used to get insights on artworks in the “Museo Novecento” in Florence,
IT. The App provides smart paradigms of interaction to ease the learning
of the Italian art history of the 20th century. Imaging Novecento exploits
automatic approaches and gamification techniques with recreational and
educational purposes. Its main goal is to reduce the cognitive effort of
users versus the complexity and the numerosity of artworks present in
the museum. To achieve this the App provides automatic artwork recog-
nition. It also uses gaming, in terms of a playful user interface which
features state-of-the-art algorithms for artistic style transfer. Automated
processes are exploited as a mean to attract visitors, approaching them
to even lesser known aspects of the history of art.
Keywords: Cultural heritage · Mobile application · Visual recognition ·

Artistic style transfer · Convolutional neural networks
1 Introduction
Modern museums can provide new paradigms for experiencing artworks. Thanks
to the technological development, novel initiatives include pervasive uses of tech
to create interactive experiences for visitors throughout a museum. However,
making content relevant and appealing through these modern technologies is a
difficult problem, requiring more and more interactivity as the audience is shift-
ing towards a ‘multimedia point of view’. Moreover, while the massive amount
of available artworks constitutes a huge resource for education and recreation
purposes, it can also be a cognitive burden for visitors.
The cognitive process related to learning has been an active subject of
study in recent decades. According to cognitive load theory, learners must cope
with a certain level of cognitive effort to process new information [22]. In this
regard, multimedia education, defined as “presenting words and pictures that

DOI: 10.1007/978-3-319-48496-9 62
782 F. Becattini et al.
are intended to foster learning” [19], can be an effective remedy because it facili-
tates the activation of sensory and cognitive perceptions (e.g. visual and notional
memory), avoiding visitors from information overloading. This can also be rein-
forced by gamification, that is the use of playful experience to help a user find
personal motivations and engagement with serious content [24]. This combina-
tion can enhance the visitor’s involvement and further lower its cognitive effort.
Using gamified applications, museum visitors have the opportunity to feel the
emotion of a game, share results with friends on social networks or become part
of a game community [20]. This aspect of learning through gaming is even more
valuable in the context of the “Bring Your Own Device” (BYOD) approach [3]
that allows on demand access to digital content on personal devices. The BYOD
approach and gamification have been identified in the NMC Horizon Report 2015
to be increasingly adopted by museums in one year’s time or less for mobile and
online engagement [15].
In this paper we report our experience in embedding these concepts into
Imaging Novecento, a system built around a mobile application developed for
the museum “Museo Novecento” in Florence, IT. We aimed at improving the
learning process of the visitors by exploiting a simple gamification paradigm,
and at reducing visitors cognitive load. To this end, we also developed a state-
of-the-art computer vision system that is able to (1) recognize artworks from
photos; (2) apply their style to user photos.
1.1 The Museo Novecento in Florence and Innovecento

The “Museo Novecento” in Florence, IT, is a museum opened on June 24, 2014.
The museum is dedicated to the Italian art of the 20th century and offers a
selection of about 300 artworks distributed in fifteen exhibition halls on two
levels. The venue is located in the former hospital of the “Leopoldine” in Piazza
Santa Maria Novella. The museum has been an example of innovation since its
genesis, thanks to the prompt adoption of the latest multimedia technologies.
In March 2015, in order to improve the visitor experience, the Municipality of
Florence has published an open call “INNOVecento - Novecento Museum Inno-
vation Lab” inviting companies and professionals to propose ideas and solutions
based on ICT. Five companies specialized in technologies applied to cultural
heritage have already responded to the call which, at the time of writing, is still
open. As NEMECH, centre of competence of the Tuscany region in Italy, we pro-
posed Imaging Novecento. The App features automatic recognition of artworks
through the visitor’s smartphone and automatic transfer of artistic styles from
artworks. These styles can be applied to user images.
1.2 Motivations and Design
The target of the App is rather wide. Although Imaging Novecento can be used
by anyone (e.g. tourists and residents), during the design process we identified a
specific audience. We mainly target the App towards people in a relatively young
Imaging NOVECENTO 783
age (between 14 and 30 years old), more accustomed to digital technologies, open
to technological innovation and to gamification.
One of the main ideas of the App is to exploit the pervasiveness of mobile
cameras in modern smartphones to reduce the cognitive effort required to
museum visitors. In fact, despite themed rooms and the ubiquitous explanatory
cards, users can still be overwhelmed by the great number of artworks present
in the museum. Labels in museums can be very concise or, on the contrary, can
be filled with lots of explanation, often generic, not highlighting salient features
of individual paintings. By using Imaging Novecento, the visitor can take a pic-
ture of the artwork he is interested in. The App will automatically recognize the
painting and provide related information. Another reason for the adoption of
this automatic process is the resistance of museums’ curators to place or attach
additional materials, such as QR codes or BLE iBeacon [12], next to artworks.
Furthermore, tourists and school groups are usually ‘hit-and-run’ visitors
who tend to rapidly forget or do not have the time to process the overload
of information. To solve this issue, Imaging Novecento leverages a playful fea-
ture that employs state-of-the-art algorithms for transferring artistic styles from
recognized artworks to user images. This is done using a gamification paradigm
at the interface level. Gamification techniques have been proved to be useful in
engaging students in the learning process, improving their skills and maximizing
their long-term memory [27].
1.3 Previous Work

Several previous works have addressed the problem of providing an engaging
experience to museum visitors. Rapid technological development has led to the
implementation of a lot of applications. There are several active trends for vir-
tual museums: immersive reality [10,18], natural interaction installations [4,9],
mixed reality, mobile applications [5,31]. While they all offer increasing engage-
ment of visitors, only recently studies on the effects of audience have been carried
out [15,21]. In particular, a recent audience study has been conducted on the
case of the “Keys to Rome” international exhibition, hosted at the “Imperial
Fora Museum” in Rome in 2015, to assess the impact of these technologies on
cultural heritage. The exhibition was made up of 11 digital installations and
applications, installed in the museum [21]. The study highlights some funda-
mental aspects that must be taken into account when designing applications for
virtual museums: (1) the majority of museum visitors are tourists and school
groups; (2) visitors generally require applications with an high level of interac-
tivity, particularly on their mobile devices; (3) it is essential for the UX design
to use metaphors of informal learning capable to stimulate attention, memory
and engagement (e.g. through gamification) in visitors.
Automatic Artwork Recognition. Automatic artwork recognition is a long stand-

ing problem in applications for cultural heritage. Descriptors such as SIFT and
SURF have been used for years in order to address this task [25,28] due to
their accuracy in recognizing paintings. Crowley and Zisserman [6] retrieve art-
works finding object correspondences between photos and paintings by using a
deformable part based method. More recent approaches for artwork recognition
adopt Convolutional Neural Networks (CNN) as in [2], where a holistic and a
part based representation are combined. Peng and Chen [23] exploit CNNs to
extract cross-layer features for artist and artistic style classification tasks. Artis-
tic style recognition is also performed in [16] on two novel large scale datasets.
Similarly to these works, we explore the use of CNNs features but we aim to
obtain a global representation that is semantically meaningful and also capable
of retaining low level visual content information. Artwork recognition has also
been used with wearable devices, as in [4] where the user’s position is jointly
estimated with what he is looking at.
Artistic Style Transfer. Regarding the application of artistic style to photos,

a lot of research has been done in the past. The problem of rendering a given
photo in the style of a particular artwork is known in literature as a branch of
non photorealistic rendering [17]. This class of works use texture transfer [7,30]
to achieve style transfer. These techniques are non-parametric and directly alter
image pixels of the content image into pre-defined styles. Another direction of
work focuses on the idea of separating style and content in order to ‘remix’
them together in different configurations. First works were evaluated on much
simpler images such as characters in different handwritings [29] or images rep-
resenting human body configurations [8]. Only recently, the breakthrough paper
from Gatys et al. [11] showed the possibility of disentangling the content from
the style of natural images by using a convolutional neural network based repre-
sentation. The advantage of this approach is the capability of performing style
transfer from any painting to any kind of content images. The approach was
recently extended with a more advanced perceptual loss [14] and also applied to
movies [1] by considering the optical flow.
2 The System
The system is composed by two main components: a mobile App and a computer
vision system responsible to address the two tasks of automatically recognize art-
works and apply artwork styles to user photos. The mobile App is used by the
visitor in the museum and is the fulcrum of the user interaction. Once installed
by the user in his mobile phone, it allows to take pictures, deliver artwork infor-
mation and request the style transfer to new photos. Due to the limited amount
of computational power available on most mobile devices, the computer vision
system is deployed on a scalable web server system that processes requests from
the mobile App. Since the two tasks use quite different technologies, we discuss
them separately in the following sections.
2.1 The Mobile App
Imaging Novecento has been developed as an Android application using Ionic1 .

Ionic is a framework, based on Sass and AngularJS, for building highly interac-
tive native web apps through mobile-optimized HTML, CSS and JS components
and tools. Imaging Novecento is a contextual App that can be used exclusively
inside the Museo Novecento in Florence. An information flyer of the App is deliv-
ered to the visitor at the ticket office. In the flyer there are a QR code, through
which the visitor can download the App from the Google Play store, and the
list of the artworks on which the App can perform the automatic recognition
and style transfer processes. The list comprises a selection of twenty artworks
for which the museum’s curators have provided multimedia materials. The App
interface (Fig. 1) is quite simple and is organized in two main views: (1) the
Camera View and (2) the Artwork Details view.
Fig. 1. Imaging Novecento in action: (1) the user takes a picture of an artwork; (2) the
artwork is recognized and insights are shown; (3) the user selects a photo from his own
gallery in order to apply that artwork style and to share the results on social networks.
The Camera View allows the visitor to frame one of the artworks on the
list in order to have it immediately recognized by the automatic system. Proper
feedback is given in case the recognition is not successful. Once the artwork is
recognized, the Artwork Details view is activated. In this view, exhaustive but
concise information about the author, the history of the artwork and its artistic
style are given. An infographic is presented to the user. It works as a “call to
action” for enabling the transfer of the recognized painting style to a photo
from the user’s device gallery. The infographic provides an animated preview
that shows the result of the artistic style transfer on a predefined picture. After
the image has been successfully uploaded, the remote process for style transfer is
performed. The result of the elaboration is then sent to the user in a few minutes
by email. The image has a resolution of 900 px wide preserving the original
image aspect ratio and can be shared on the most popular social networks (e.g.
Facebook).
1
http://ionicframework.com/.
2.2 Automatic Artwork Recognition
Artwork recognition is performed through a Python web server with a REST

interface. The server processes the image and returns the ID of the recognized
painting. The recognition step combines modern deep features with classical
Support Vector Machines (SVM) in order to classify photos of paintings. Image
features are extracted using a deep convolutional neural network (CNN), and
are then evaluated using a set of classifiers, one for each recognizable artwork.
The neural network we adopted is the Caffe reference model [13], fine-tuned for
style recognition using the FlickrStyle dataset [16]2 .
In order to obtain a representation which is at the same time semantically
meaningful and capable of retaining low level visual content information, we
extract image features from an intermediate level of the network. In particu-
lar, we adopt the pool5 feature map, the latest one before the fully connected
(FC) layers of the CNN. In fact, FC layers trade spatial information for a more
semantic representation, which is highly coupled with the task and with the
visual domain on which the network has been trained. This choice is therefore
motivated by the fact that our visual domain, while being quite close, is different
from the one of FlickrStyle. Moreover, since a sufficiently large dataset was not
available to perform a further fine-tuning step, SVM classifiers have been trained
to adapt the framework to the App’s domain and be able to classify artworks
correctly. For training the classifiers we used approximately 1,800 images, gath-
ered at the museum using different smartphones and tablets, namely Galaxy S4,
Galaxy Tab, iPhone 6, iPad Mini and OnePlus One. These images represent all
of the twenty artworks plus a ‘negative’ set of images containing other scenes
and paintings inside of the museum. They are used to reduce the false positive
rate when the user accidentally attempts to recognize other paintings. All the
classifiers are One vs All SVMs. During the evaluation phase, the ID of the high-
est scoring one is returned to the mobile App, if it scores above a cross validated
threshold. Details about the recognized artwork are then provided to the user,
who can upload a personal photo to get the style of the painting transferred on
to it. Calls to the webserver are handled asynchronously and each request takes
approximately 300 ms on a CPU.
In order to test the recognition accuracy we collected an additional set of pho-
tos which were not used for training. For each one of the twenty artworks in our
system, we collected approximately 30 photos taken from different viewpoints,
with different scales and degrees of occlusion. Figure 2 shows some of the photos
from the test set. Some of them are “difficult” in a sense that might be blurred
or taken from challenging viewpoints and artworks may be partially occluded
by other visitors. Despite these difficulties our system achieves an overall good
performance with a mean accuracy of 94.01 %. In detail, in Fig. 3 we report the
confusion matrix for the twenty artworks in the test set, showing how often each
painting is correctly classified or confused with other artworks. As can be seen,
2
The network is available online at http://caffe.berkeleyvision.org/gathered/
examples/finetune flickr style.html.
Fig. 2. Samples from the dataset collected at the museum. On the first row standard
pictures are shown, depicting the painting in their entirety. On the second row instead,
are reported more challenging photos, due to blur, occlusion or rotation.
Fig. 3. Confusion matrix for the artwork recognition module. Each row indicates the
percentages of correct and incorrect classifications for a given artwork.
the majority of artworks are perfectly recognized. Only four artworks have per-
formance slightly inferior to 0.9, due to the difficult lightening conditions present
in their specific locations at the museum.
2.3 Artistic Style Transfer

From the Artwork Detail view of the mobile App, the user has the possibility
to upload a personal image on which the style of the artwork will be applied.
In this way, entertaining personal pictures that share similarities with the art-
works can be obtained and shared on social networks. As a result, a visit at
the museum can become a playful experience, combining gaming and learning
aspects for young visitors. We base our approach on that of Gatys et al. [11],
that is capable of freely mixing style and content of two different photos. The
main advantage of this approach is its broad applicability to different styles,
in contrast to fixed handcrafted styles [7,30]. This allows a museum curator to
easily add new artworks in the system without requiring the development of a
new transfer style algorithm. Following [11], our approach uses a CNN to derive
a neural representation of content and style. The feature responses of a pre-
trained network on object recognition (VGG-19 [26]) are used to capture the
appearance of an artwork image and the content of a user photo under the form
of texture information. We start from a blank novel image that is altered with
back-propagation until its neural representation is similar in terms of euclidean
distance to the style and content representations (Fig. 4).
Fig. 4. Two examples of image stylization: (1) Baccio Maria Bacci, “Il tram di Fiesole”,
applied to a picture of the Battistero in Florence, IT; (2) Alberto Moretti, “Malcom X
ed altri”, applied to a picture of Piazza della Repubblica, also in Florence.
Unfortunately, the generation of the image is quite computational intensive.

For an image of 900 pixel large, it takes about ∼90 seconds on a K80 NVIDIA
GPU. As a result, the requests have to be handled offline since it is not possi-
bile to obtain the output image in few seconds. Considering also that multiple
requests can be made at the same time from multiple users, we implemented a
scalable web server that is able to be easily deployed on several interconnected
nodes. Web requests are handled in Python and enqueued to a distributed queue
run by a Celery3 server. By treating each request as a single unit task, it allows
to process the images in a distributed batch fashion on several GPUs and several
servers if available. After completing the computation, each output image is sent
to the user via email, together with a description of the artwork. We also include
links to share the image to several social media, with the aim of enabling viral
publicity of the museum.
3 Conclusion
We presented the Imaging Novecento App, recently developed for the “Museo
Novecento” in Florence, IT. Following previous studies on cultural heritage audi-
ence and applications, the App aims at enhancing the experience in the museum
reducing cognitive load and exploiting gamification. The App automatically
recognizes a selection of paintings and provides insights on artworks and their
authors. The user can upload a personal picture with his smartphone to get it
stylized with the recognized artwork style. He also has the possibility of sharing
it on social networks. In the paper we show how computer vision technologies
can be exploited to increase interactivity and reduce cognitive load. This can
attract the targeted audience to the museum and further engage people with
content.
Acknowledgments. We acknowledge the support of the “Museo Novecento” in

Florence, IT, and the Municipality of Florence, IT. This research was supported
by “THE SOCIAL MUSEUM AND SMART TOURISM”, MIUR project no.
CTN01 00034 23154 SMST.
References
1. Anderson, A.G., Berg, C.P., Mossing, D.P., Olshausen, B.A.: Deepmovie:
using optical flow and deep neural networks to stylize movies. arXiv preprint
arXiv:1605.08153 (2016)
2. Anwer, R.M., Khan, F.S., van de Weijer, J., Laaksonen, J.: Combining holistic
and part-based deep representations for computational painting categorization. In:
Proceedings of the 2016 ACM ICMR, pp. 339–342. ACM (2016)
3. Ballagas, R., Rohs, M., Sheridan, J.G., Borchers, J.: Byod: bring your own device.
In: Proceedings of the Workshop on Ubiquitous Display Environments, Ubicomp,
vol. 2004 (2004)
4. Baraldi, L., Paci, F., Serra, G., Benini, L., Cucchiara, R.: Gesture recognition
using wearable vision sensors to enhance visitors Museum experiences. IEEE Sens.
J. 15(5), 2705–2714 (2015)
5. Chen, C.Y., Chang, B.R., Huang, P.S.: Multimedia augmented reality information
system for Museum guidance. Pers. Ubiquit. Comput. 18(2), 315–322 (2014)
6. Crowley, E., Zisserman, A.: The state of the art: object retrieval in paintings using
discriminative regions. In: BMVC (2014)
3
http://www.celeryproject.org.
7. Efros, A.A., Freeman, W.T.: Image quilting for texture synthesis and transfer. In:
Proceedings of the 28th Annual Conference on Computer Graphics and Interactive
Techniques, pp. 341–346. ACM (2001)
8. Elgammal, A., Lee, C.S.: Separating style and content on a nonlinear manifold. In:
Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision
and Pattern Recognition, CVPR 2004, vol. 1, pp. I–478. IEEE (2004)
9. Fanini, B., d’Annibale, E., Demetrescu, E., Ferdani, D., Pagano, A.: Engaging and
shared gesture-based interaction for museums the case study of K2R international
expo in rome. In: 2015 Digital Heritage, vol. 1, pp. 263–270. IEEE (2015)
10. Fineschi, A., Pozzebon, A.: A 3D virtual tour of the Santa Maria della Scala
Museum Complex in Siena, Italy, based on the use of Oculus RIFT HMD. In: 2015
International Conference on 3D Imaging (IC3D), pp. 1–5. IEEE (2015)
11. Gatys, L.A., Ecker, A.S., Bethge, M.: A neural algorithm of artistic style. arXiv
preprint arXiv:1508.06576 (2015)
12. He, Z., Cui, B., Zhou, W., Yokoi, S.: A proposal of interaction system between
visitor and collection in Museum hall by iBeacon. In: 2015 10th International Con-
ference on Computer Science & Education (ICCSE), pp. 427–430. IEEE (2015)
13. Jia, Y., Shelhamer, E., Donahue, J., Karayev, S., Long, J., Girshick, R., Guadar-
rama, S., Darrell, T.: Caffe: Convolutional architecture for fast feature embedding.
In: Proceedings of the 22nd ACM MM, pp. 675–678. ACM (2014)
14. Johnson, J., Alahi, A., Fei-Fei, L.: Perceptual losses for real-time style transfer and
super-resolution. arXiv preprint arXiv:1603.08155 (2016)
15. Johnson, L., Adams Becker, S., Estrada, V., Freeman, A.: The NMC Horizon
Report: 2015 Museum Edition. ERIC (2015)
16. Karayev, S., Trentacoste, M., Han, H., Agarwala, A., Darrell, T., Hertzmann, A.,
Winnemoeller, H.: Recognizing image style. arXiv preprint arXiv:1311.3715 (2013)
17. Kyprianidis, J.E., Collomosse, J., Wang, T., Isenberg, T.: State of the “Art”: a
taxonomy of artistic stylization techniques for images and video. IEEE Trans. Vis.
Comput. Graph. 19(5), 866–885 (2013)
18. Malomo, L., Banterle, F., Pingi, P., Gabellone, F., Scopigno, R.: Virtualtour: a
system for exploring cultural heritage sites in an immersive way. In: 2015 Digital
Heritage, vol. 1, pp. 309–312. IEEE (2015)
19. Mayer, R.E., Moreno, R.: Nine ways to reduce cognitive load in multimedia learn-
ing. Educ. Psychol. 38(1), 43–52 (2003)
20. Negruşa, A.L., Toader, V., Sofică, A., Tutunea, M.F., Rus, R.V.: Exploring gam-
ification techniques and applications for sustainable tourism. Sustainability 7(8),
11160–11189 (2015)
21. Pagano, A., Armone, G., De Sanctis, E.: Virtual Museums and audience studies:
the case of “Keys to Rome” exhibition. In: 2015 Digital Heritage, vol. 1, pp. 373–
376. IEEE (2015)
22. Park, S.: The effects of social cue principles on cognitive load, situational interest,
motivation, and achievement in pedagogical agent multimedia learning. J. Educ.
Technol. Soc. 4, 211–229 (2015)
23. Peng, K.C., Chen, T.: Cross-layer features in convolutional neural networks for
generic classification tasks. In: 2015 IEEE International Conference on Image
Processing (ICIP), pp. 3057–3061. IEEE (2015)
24. Reiners, T., Wood, L.C. (eds.): Gamification in Education and Business. Springer
International Publishing, Switzerland (2015)
25. Ruf, B., Kokiopoulou, E., Detyniecki, M.: Mobile Museum guide based on fast
SIFT recognition. In: Detyniecki, M., Leiner, U., Nürnberger, A. (eds.) AMR
2008. LNCS, vol. 5811, pp. 170–183. Springer, Heidelberg (2010). doi:10.1007/
978-3-642-14758-6 14
26. Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale
image recognition. arXiv preprint arXiv:1409.1556 (2014)
27. de Sousa Borges, S., Durelli, V.H., Reis, H.M., Isotani, S.: A systematic mapping
on gamification applied to education. In: Proceedings of the 29th Annual ACM
Symposium on Applied Computing, pp. 216–222. ACM (2014)
28. Temmermans, F., Jansen, B., Deklerck, R., Schelkens, P., Cornelis, J.: The mobile
Museum guide: artwork recognition with eigenpaintings and SURF. In: WIAMIS
2011, Delft, The Netherlands, April 13–15, 2011. TU Delft; EWI; MM; PRB (2011)
29. Tenenbaum, J.B., Freeman, W.T.: Separating style and content with bilinear mod-
els. Neural Comput. 12(6), 1247–1283 (2000)
30. Xie, X., Tian, F., Seah, H.S.: Feature guided texture synthesis (FGTS) for artis-
tic style transfer. In: Proceedings of the 2nd International Conference on Digital
Interactive Media in Entertainment and Arts, pp. 44–49. ACM (2007)
31. Yoon, S.A., Wang, J.: Making the invisible visible in science museums through
augmented reality devices. TechTrends 58(1), 49–55 (2014)
Towards the Design of a User-Friendly and Trustworthy
Mobile System for Museums
Kostas Koukoulis and Dimitrios Koukopoulos ✉

( )
Department of Cultural Heritage Management and New Technologies, University of Patras,

Agrinio, Greece
{kkoukoulis,dkoukopoulos}@upatras.gr
Abstract. Designing mobile applications for enhancing user visiting experi‐

ences in museums is a current trend. This paper discusses the current work on
mobile applications that are dedicated to museums. Following a specific meth‐
odology, we propose specific user and app classification requirements. Based on
those requirements, we present the design of a user-friendly and trustworthy
mobile system prototype that includes functionality needed from such institu‐
tions. We adopt suitable authorization mechanisms permitting specific operations
to various user groups. We evaluate the proposed system design comparing it with
other known systems following a specific usage scenario. We feel that this study
could help on the design and implementation of trustworthy mobile-based
museum visiting applications.
Keywords: Cultural heritage · Museum mobile applications · Digitization ·

Multimedia databases · Web smartphone-based services
1 Introduction
Museums are in a rush to follow the trend of our era. According to the Museums Asso‐
ciation [1] they started providing specialized mobile applications to the users of the
museums respecting the fact that the use of mobile smartphones has exploded over the
past years. This trend is being extensively adopted by large institutions for the profound
reason that they have the funds to finance such a change. Their intensions are to offer
extra content to the visitors, to make them more energetic during their tour, to attract
new visitors, to expand the access to people with special needs and finally to make
visitors participants of their activities [1]. Ambrose and Paine [20] insist that a museum
should know its users and should have answers on questions such as: why will people
visit my museum, who will visit, how much time they will spend etc. The fact is that
today museum visitors, like everybody else, use a smartphone. But the adaption of tech‐
nology should not be made blindly. Some skeptics think that the museum transforms in
the base of database design [13]. Others support the use of technology as a need for the
new generation [14].
An important challenge in cultural web-based systems is long-term availability that
includes digital rights management and content preservation from unauthorized access

DOI: 10.1007/978-3-319-48496-9_63
Towards the Design of a User-Friendly and Trustworthy Mobile 793
[28]. This problem is even worse in mobile systems that support management, collection
and dissemination of multimedia content in cultural environments due to the inherent
characteristics of mobile devices (operating systems gaps and absence of access control
and data security mechanisms) [27, 29].
In this work, we demonstrate a methodology of designing and evaluating a mobile
system for cultural institutions, like museums. The goals of our approach are two-fold:
enhancing visit experience for end-users and facilitate management and preservation
actions for the institution personnel concerning museum visits. Thus, we investigate user
requirements and mobile application capabilities that can be exploited for the design of
a friendly, useful and trustworthy mobile system serving visiting and management
purposes. We analyze the basic system usage scenario targeting on the study of specific
user groups (personnel or visitors), specifying cultural content types, attributes and
relations, limiting permitted user operations on cultural content for security purposes
and supporting specific services that can be applied before, during and after a museum
visit. Having these potentials in mind, we propose a system architecture for the trust‐
worthy and user-friendly adaptation of mobile services into the everyday life of a
museum. Furthermore, we investigate operationally the proposed system in order to
understand the communication interactions and needs of the various entities partici‐
pating in system’s design. Finally, we make an attempt to evaluate our system design
involving specific experts in museum technologies and environments. In this attempt,
we compare the proposed system to other well-known museum visiting applications [3,
4, 6] following s specific simple usage scenario.
2 Related Works
In 2006 there was a big challenge for the digitalization process [15], the majority of the
museums already had web sites and also used other technologies such as accounting
software, broadband internet connection, computerized catalogs, computerized collec‐
tions management systems etc. Nowadays a museum includes not only a website, but
also a social media presence, mobile tools and apps, electronic fundraising [16], and
much more such as some management system for its assets [17, 18]. The Acropolis
Museum [12] has released digital applications for home users regarding the exhibits,
has created a digital repository for educational purposes and provides on site visual
material in portable electronic media. The Smithsonian Institute [19] provides applica‐
tions for educational purposes and a vast amount of material is digitized, categorized
and provided for researchers, kids and educators. A large number of apps exist already
in Google’s play store and the majority is being built by independent companies
containing specialized material [21, 22]. A small amount of them were built by the
institutions but the documentation is unavailable to the public. The British Museum Lite
[23] allows a virtual tour at British Museum and provides maps, high resolution images
and thematic catalogues of exhibits. Smithsonian Institution [24] provides its own app.
It contains visit planning, search on museums assets, AR and information about the
museum. The full version costs a few dollars and provides full functionality. A lot of
research has been done in the field of authorization mechanisms concerning access
794 K. Koukoulis and D. Koukopoulos
control in multimedia content [25–27]. An authorization model specialized for multi‐

media content stored in digital libraries is described in [25]. Content-based access
control is applied on multiple files of digital art projects in a collaborative web-based
environment [26]. An authorization mechanism based on role-based access control is
applied on smartphone devices for protecting multimedia content which is disseminated
though multimedia guiding services in a cultural environment in [27]. Considering the
fact that Museum Mobile Apps are being built for such a short period, evaluation
methods are not yet published. Some work has been done in the past few years in
designing mobile applications for cultural institutions. Chess project makes efforts to
use AR, digital storytelling and personalized tour with the use of personas [3]. Another
implementation made by researches in Korea proposed a system that provides a unified
presentation of the content of the museums in a whole region [6]. Other researchers
proposed a system that implements a tour based on the knowledge of other people
providing again functionality such as AR [4].
3 User Requirements and App Classification
Till now there are a few works that deal with the investigation of user requirements for
applications offering services in environments of cultural interest (museums, archaeo‐
logical sites, city centers, etc.) [5, 7, 8]. Hui-Ying and Chao-Chien [5] talked about
visitor’s expectations of a visit to an archeological site and this work was the base for
structuring our user requirements. Palumpo et al. [7] started a study to identify the key
factors related to visitor satisfaction. They ended up with six must have functionalities.
Along with these requirements, there were placed data from a Smithsonian’s [8] wide
research that aimed to investigate their visitors needs regarding to their visits at the
institution. Analyzing all these data we ended up with a set of axes that a mobile appli‐
cation should implement. Based on the Smartphone functionalities (camera, micro‐
phone, Wi-Fi connection, Bluetooth, touch screen capabilities, GPS sensor) and inter‐
views taken from museum experts we propose a number of user requirements for a
mobile application (Fig. 1).
Fig. 1. User requirements

Fig. 2. Basic tables in composite DB
Fig. 3. Services
Fig. 4. Simple usage scenario
Some work has been done in order to classify the applications regarding their features
[2]. Their classification includes axes such as general characteristics (free or paid app),
type of the application (guided tours, devoted to specific artworks, games, content crea‐
tion by user), use of the app (during, before or after their visit) and functionality (contains
web 2.0, AR, various views). Data was also taken from Smithsonian’s research [8]. App
classification parameters based on technological characteristics that specialize the appli‐
cation environment can be classified in three categories: (i) technology and features (like
installations on all mobile environments, use network capabilities and mobile sensors,
AR, interaction and personalization), (ii) security (like user different rights and func‐
tionalities, existence of security mechanism capabilities or not), and (iii) app reliability
concerning the computing environment (like easy downloading of the app, connection
to museums network and free downloading).
4 Usage Scenario
In the process to define usage scenarios for a proposed system we need to declare the
role of its users, the type of content that the application will manage and diffuse the
operations that will be permitted on data and usage scenarios.
There are six types of users able to connect to the app and have different functionality:
curators, specialists, administrators, guests, visitors and students. Each of these groups
of users should be able to access the corresponding app screens, data and functionality.
Users could be grouped in two basic categories project team members (curators, special‐
ists, administrators) and internet users (visitors, guests, students). Curators: Should be
able to have full access of databases and produce the composite data. They also choose
the specialists team and decide with the director what can be digitized. Specialists:
Should be able to manage content data of the database through app screens, to alter
database data, to add new exhibitions, to create tour routes and to add information about
the museum. Their contact info should be also in the database. In real, they are artists,
creators or scientists. Administrators: Provide to other types of users the corresponding
screens and maintain the database and the hardware that supports the app. Visitors:
Museum visitors or just the users of the app. Some researchers indicate that the users’
needs are relevant to their background [8]. These needs differentiate according to their
place of residence, frequency of visits, use of museum aids, reasons for using their
phones and their styles of accessing information. Guests: They can access limited func‐
tionality and data in order to have a clear view for using the proposed app. Students:
Could access all the functionality assigned to visitors and also view primitive data,
specialist’s comments, contact info of specialists etc. Guests and students are a subtype
of visitor.
The app content refers to all kind of data that will be held in the database concerning
places of the museum and users. Images: Images of exhibits, collections, museum halls
and its exteriors, architectural drawings, floor and hall plans containing the exact position
of the exhibits and city area maps. Audio: not a main exhibition item, but a way to provide
the visitor information about an exhibit and to facilitate persons with disabilities in their
tour in an advanced version of the system. Video: from excavations, maintenance,
exhibit transportations to other museums or from any special time of the course of the
exhibit. Video of specialist interviews, visitor’s experiences, future visitor’s comments
and virtual tours. Text: in-depth object info, descriptions of exhibits and collections,
specialist interview transcripts, museum info and user’s comments. All entities included
in the database should have a text. GPS/routing data: outdoor routing data like positions
and routes of exhibits, places of interest, public transportation stations and data for
indoor routing using Bluetooth and beacons [9].
All content should be classified in a number of main categories and should be placed
in separate databases. Primitive museum data (Primitive DB): Specialists provide data
regarding the exhibits and collections (images, texts, sound, video and multimedia).
Composite museum data (Composite DB): Produced by primitive data and data provided
by visitors (comments, photos, videos) and guests like personalized tours content. Visi‐
tors and students material along with the recorded path of their visit to the museum could
constitute the basis for a collection of stories referring to different user interests. User
credentials (User Credential DB): Login information and user preferences. Logistic data
(Logistics DB): Information regarding the ticketing service, way finding, security serv‐
ices, hours of operation, parking or metro access, restrooms and crowd information (data
provided by special trained personnel). User-generated content (User Content DB):
Data to be stored by internet users. These data are kept separately from the composite
data but are connected with the corresponding exhibition id. Security Data (Security
DB): Any access to the database will be recorded.
Data are kept in separate databases as shown in Fig. 5. The basic work that is made
on the data could be divided in a number of categories: insert, delete, view, search,
annotate and relate. Specialists have full access to their own data on primitive content
database. Curators may also annotate these data and are responsible for creating compo‐
site data on a distinct database. Content created by internet users is placed on a different
database. The administrator of the system does not manipulate data regarding museum
assets. He should only be able to assign privileges to users and to keep the system
running, doing whatever is needed from the technological side.
Fig. 5. System architecture
For internet users the main action is viewing. The app should give them the option
to choose among different kinds of views. For example, they should be able to view a
full time video or a portion of it or just a scene. Some of the services devoted to visitors
are a part of the dissemination. The system should contain games related to the museum
content, give visitors the right to publish some of the content to social networks such as
recorded visits structuring their own scenarios like story-telling process or even to give
them the functionality to buy goods from the museum store. Internet users are able to
create their own material and their data should be stored in a separate DB following the
schema of “Basic tables in composite DB” (Fig. 2).
A guest should be able to view the smaller amount of information. He may become
a visitor and the provided information needs to meet that need. He should be able to
view informational data about the museum such as opening hours and navigation to the
museum, data regarding to the museum exhibits and collections in order to decide if he
wants to visit it and to just gain some knowledge. The only insert that he could do on
data is some rating. Visitors should be able to access and view data regarding the exhibits
and the collections, to see personalized routes to follow on their visit, to hear audio
descriptions, use augmented reality etc. Also, they can input data taken during their visit
such as photographs and put their comments and ratings. A more demanding visitor is
the type of student. Students have the same rights as visitors plus the option to view
scientific papers and specialist comments and data. Students can store locally certain
data that the curator permits. The main objective of the app is to provide them with
material for their studies and researches. Permissions for accessing this material should
be given to them for a specific data set from the curator.
The main issue was to assign privileges to internet users. The data manipulation is
made through the app’s services. Each user type is assigned to a number of screens in
his app and de facto with the corresponding services. In the assigned screens there was
a need to show or hide data. The publication right field on podcast, video, images and
text contains the rights for viewing the corresponding data for each subtype of internet
users. The estimated guide time is a model based characteristic. It is needed in order to
automatically calculate tours on demand based on visitor’s available time.
5 Services and Scenarios
Our goal was to build an application that will contain all the necessary features that will
help the visitors of a cultural institution to have an integrated tour experience. We created
a questionnaire in order to determine the basic needs of a simple user from the perspec‐
tive of museum experts. According to the answers that we had, we classify application
services for the internet user in three main categories: pre-touring, in-touring and post-
touring operations (Fig. 3).
A simple scenario is created for the evaluation of the proposed app. This scenario
contains basic actions that a visitor usually performs visiting a museum (Fig. 4).
6 Proposed System Architecture
The proposed mobile app delivers services for both internet users and project team
members. An application connected to the database is also used for ticketing and shop‐
ping services by museum staff. The system is separated in modules. The first module of
the system contains a database set (Security DB, Primitive DB, User credentials DB,
User content DB, Composite DB) and a broker in order to facilitate their use (Database
management module). Data will be accessed only through services and the direct data‐
base will be prohibited. Any access to the main application data will be recorded by the
security database. The second module is the Application Server where is installed all
the logic of the proposed system in order to access our data with safety. Complementary
the application server can support commercial transactions between users and museum,
like user ticketing and shopping that are necessary functionalities in modern museums
for funding purposes. The third module contains Mobile Services and its role is to
interact successfully and efficiently with the users. Also, users through the app can make
commercial transactions with museum. The proposed system is implemented as a three-

tire application. For the data layer we use a DBMS like SQL Server. The application
layer uses an application server (such as Microsoft’s IIS). The presentation layer
contains the mobile app which operates in Android smartphones that use SDK 24.4.1
and systems for ticketing and shopping that are based on HTML5 and angularJS.
7 Operational Design Scenarios
Search for an exhibit by name: The visitor stands before the exhibit and read its label.
His main objective is to find information regarding the specific item. The application
has simple search functionality for cases like this. He touches the menu and tap on the
search icon. Immediately the screen gets blare and a single line search input field appears
on the top. The visitor enters the exhibit name. The mobile app passes the search keyword
to the corresponding functional unit which in its turn request information from the
broker. The broker routes the question to one or more databases and returns a list of
items to the functional unit that is provided to the visitor.
Visitor chooses to see what is contained in adjoining rooms: The visitor is in the
museum and located to a hall. He doesn’t want to be in that specific area anymore and
wants to have an overview of the halls next to him. He taps on the apps menu and chooses
“What’s next to me”. Immediately the action goes to the Application server along with
his position data. Broker chooses the composite database and asks for hall plans of places
next to his position. These plans are being forwarded to the Application Server who
sends them to the app.
Following a recommended route based on the time available and personal interests:
The visitor finds out that he hasn’t got much time. He decides to let the system propose
a route defining the available time and his interests. From his screen he chooses “Give
me a proposed tour”. Then the app asks him to fill in information about his interests (like
era, artists, value of the exhibits). Visitor has one hour left and wants to see the most
popular exhibits. Then a request is being made to the application server. The application
server does a query to databases asking information about popular exhibits (rated by
visitors). For each exhibit the application server asks databases for the estimated time
needed for viewing it. Subsequently, it creates a route considering his position and the
criteria mentioned above. The route is provided to the visitor.
8 Evaluation
In order to evaluate the proposed system, we adopt the method proposed by Lewis et al.
known as cognitive walkthrough [10]. The basic advantage of this method relies on the
fact that there is no need of fully implementing the system you need to evaluate [11].
There are two phases in this method: preparation and evaluation. In the preparation phase
there is a need to determine a set of representative tasks that the application is intended
to support. We selected the tasks of the simple usage scenario on Fig. 4. In the second
phase, a small number of specialists from fields related to museum technologies and
environments, read the application specifications and grade each user action. Grades
varied from 0 to 4. Zero indicates the absence of the service. The grades 1 to 4 correspond
to the familiarity of users towards specific acts (the percentage of users who know to
use a system service). Four is the maximum degree and 1 is the minimum one. In our
evaluation, we compare the proposed system with systems presented in [3, 4, 6]. The
proposed system seems to get the highest score (Fig. 6).
Fig. 6. Walkthrough evaluation
9 Conclusions
The proposed system attempts to satisfy the needs of users, experts or not, concerning
the visit experience in a museum. It exploits the functionalities of modern smartphone
devices in order to offer efficient visiting services in a friendly manner. We have taken
into account user requirements (and app classifications requirements), usage scenarios
(users, content classification and permitted content operations) to build services and
scenarios that would be served by the proposed system. After system design, we started
to implement the proposed system services in order to support the basic usage scenario.
A first primitive evaluation indicates success on the design of the proposed system and
the produced comparison validates it. Following we are going to implement the full
system functionality targeting specific museum environments (like general, technology,
history and art museums [30]) and distribute our mobile app to the broad community
through the App Store. In this process, we are going to ask user community for contin‐
uous feedback in terms of ease of use, potential usefulness, perceived usefulness and
security of use.
Acknowledgement. This work is an internal project of Cultural Technology Laboratory of the

Department of Cultural Heritage Management and New Technologies of the University of Patras
having as target group the local museums and cultural heritage environments of Aitoloakarnania,
a prefecture in Greece.
References
1. Atkinson, R.: 2013 MA mobile survey. Technical report, Museums Association (2013)
2. Economou, M., Meintani, E.: Promising beginning? evaluating museum mobile phone apps.
In: Ciolfi, L., Scott, K., Barbieri, S. (eds.) Rethinking Technology in Museums 2011:
Emerging Experiences, pp. 87–101. University of Limerick, Ireland (2011)
3. Pujol, L., Katifori, A., Vayanou, M., Roussou, M., Karvounis, M., Kyriakidi, M., Eleftheratou,
S., Ioannidis, Y.: From personalization to adaptivity–creating immersive visits through
interactive digital storytelling at the Acropolis Museum. In: Botia, J.A., Charitos, D. (eds.) IE
2013, pp. 541–554. IOS Press, Athens (2013)
4. Huang, W., Kaminski, B., Luo, J., Huang, X., Li, J., Ross, A., Wright, J., An, D.: SMART:
design and evaluation of a collaborative museum visiting application. In: Luo, Y. (ed.) CDVE
2015. LNCS, vol. 9320, pp. 57–64. Springer, Heidelberg (2015). doi:10.1007/978-3-
319-24132-6_7
5. Shi, H.Y., Chen, C.C.: A study of service quality and satisfaction for museums-taking the
national museum of prehistory as an example. J. Hum. Res. Ad. Learn. 4(1), 159–170 (2008)
6. Bae, E.S., Im, D.U., Lee, S.Y.: Smart museum based on regional unified app. Int. J. Soft. Eng.
Appl. 7(4), 157–166 (2013)
7. Palumbo, F., Dominici, G., Basile, F.: Designing a mobile app for museums according to the
drivers of visitor satisfaction. In: Raguz, I.V., Roushdy, M. (eds.) Recent Advances in
Business Management and Marketing, pp. 159–166. WSEAS Press, USA (2013)
8. Smart phone services for Smithsonian visitors. Technical report, Smithsonian Institution
(2010)
9. Chawathe, S.: Beacon placement for indoor localization using bluetooth. In: 11th International
IEEE Conference on Intelligent Transportation Systems, pp. 980–985. IEEE Press, Beijing
(2008)
10. Lewis, C., Polson, P., Wharton, C., Rieman, J.: Testing a walkthrough methodology for
theory-based design of walk-up-and-use interfaces. In: CHI 1990, pp. 235–242. ACM Press,
New York (1990)
11. Polson, P., Lewis, C., Rieman, J., Wharton, C.: Cognitive walkthroughs: a method for theory-
based evaluation of user interfaces. J. Man Mach. Stud. 36(5), 741–773 (1992)
12. Learning resources, Acropolis Museum. http://www.theacropolismuseum.gr
13. Pepi, M.: Is a museum a database?: institutional conditions in net utopia. e-flux 60 (2014)
14. Connecting museums with generation Z. Technical report, Smack Agency (2013)
15. Status of technology and digitization in the nations’ s museums and libraries. Technical report,
Institute of Museum and Library Services (2006)
16. Johnson, L., Adams Becker, S., Estrada, V., Freeman, A.: NMC horizon report: 2015 museum
edition. Technical report, The New Media Consortium (2015)
17. Koukopoulos, D., Tsolis, D., Heliades, G.: Ionian music archive: application of digitization,
management and dissemination technologies for musical cultural heritage. In: IISA 2014, pp.
239–244. IEEE Press, Chania (2014)
18. Chaplin, E., Tullock, J.: Successfully managing archives in museums. Technical report, AIM
Success Guides (2015)
19. Smithsonian. https://www.si.edu
20. Ambrose, T., Paine, C.: Museum Basics, 3rd edn. Routledge, Oxford (1993)
21. Museums in NYC. https://play.google.com
22. Rodin museum. https://play.google.com
23. The British Museum Lite. https://play.google.com
24. Smithsonian mobile. https://play.google.com
25. Kodali, N., Farkas, C., Wijesekera, D.: An authorization model for multimedia digital
libraries. J. Dig. Libr. 4(3), 139–155 (2004)
26. Koukopoulos, D., Styliaras, G.: Security in collaborative multimedia web-based art projects.
J. Mult. 5(5), 404–416 (2010)
27. Koukopoulos, D., Styliaras, G.: Design of trustworthy smartphone-based multimedia services
in cultural environments. J. Electron. Commun. Res. 13(2), 129–150 (2013)
28. Arnold, D.: Paster’s quadrant: cultural heritage as inspiration for basic research in computer
science. J. Comput. Cult. Her. 2(1), 1–10 (2008)
29. Oberheide, J., Jahanian, F.: When mobile is harder than fixed (and vice versa): demystifying
security challenges in mobile environments. In: HotMobile 2010, pp. 43–48. ACM Press, New
York (2010)
30. Types of Museum. http://www.britannica.com/topic/types-of-museums-398830
Project Paper: Serious Games for
Cultural Heritage
Project iMARECULTURE: Advanced VR, iMmersive
Serious Games and Augmented REality as Tools to Raise
Awareness and Access to European Underwater
CULTURal heritagE
D. Skarlatos1 ✉ , P. Agrafiotis1,2, T. Balogh3, F. Bruno4, F. Castro5,

( )
B. Davidde Petriaggi , S. Demesticha7, A. Doulamis2, P. Drap8, A. Georgopoulos2,

6
F. Kikillos12, P. Kyriakidis1, F. Liarokapis9, C. Poullis10, and S. Rizvic11

1
Cyprus University of Technology, Limassol, Cyprus
[email protected]
2
National Technical University of Athens, Athens, Greece
3
Holografika Hologrameloallito Fejleszto Es Forgalmazo Kft, Budapest, Hungary
4
3D Research s.r.l., University of Calabria, Rende, Italy
5
Universidade Nova de Lisboa, Lisbon, Portugal
6
Ministero dei Beni e delle Attività Culturali e del Turismo, ISCR, Rome, Italy
7
University of Cyprus, Nicosia, Cyprus
8
Aix-Marseille University, Marseille, France
9
Masaryk University, Brno, Czech Republic
10
Concordia University, Montreal, Canada
11
Sarajevo University, Sarajevo, Bosnia and Herzegovina
12
Pierides Foundation, Larnaca, Cyprus
Abstract. The project iMARECULTURE is focusing in raising European iden‐

tity awareness using maritime and underwater cultural interaction and exchange
in Mediterranean Sea. Commercial ship routes joining Europe with other cultures
are vivid examples of cultural interaction, while shipwrecks and submerged sites,
unreachable to wide public are excellent samples that can benefit from immersive
technologies, augmented and virtual reality. The projects aim to bring inherently
unreachable underwater cultural heritage within digital reach of the wide public
using virtual visits and immersive technologies. Apart from reusing existing 3D
data of underwater shipwrecks and sites, with respect to ethics, rights and
licensing, to provide a personalized dry visit to a museum visitor or augmented
reality to the diver, it also emphasizes on developing pre- and after- encounter of
the digital or physical museum visitor. The former one is implemented exploiting
geospatial enabled technologies for developing a serious game of sailing over
ancient Mediterranean and the latter for an underwater shipwreck excavation
game. Both games are realized thought social media, in order to facilitate infor‐
mation exchange among users. The project supports dry visits providing immer‐
sive experience through VR Cave and 3D info kiosks on museums or through the
web. Additionally, aims to significantly enhance the experience of the diver,
visitor or scholar, using underwater augmented reality in a tablet and an under‐
water housing. The consortium is composed by universities and SMEs with expe‐
rience in diverse underwater projects, existing digital libraries, and people many
of which are divers themselves.

DOI: 10.1007/978-3-319-48496-9_64
806 D. Skarlatos et al.
Keywords: Underwater · Archaeological sites · Shipwrecks · Maritime · Virtual

museums · Serious games · Immersive · Holography · European identity
1 Introduction
The area of Virtual Museums, Virtual Guides and Virtual Reconstruction of Cultural
Heritage, has a number of past and currently important active projects of this scope (e.g.
V-MUST, F-MU.S.EU.M., VENUS, MINERVA, MINERVA PLUS, MINERVA EC,
THE MICHAEL PLUS, ATHENA, ATHENA PLUS, ARCHEOGUIDE, 3DMURALE
etc.). However, these projects do not address the real challenge of an Underwater Virtual
Museum. In addition, projects related to underwater cultural heritage and environments
are not engaged with the challenge of Virtual Museums and Immersive Technologies.
These projects (e.g. SASMAP, WRECKPROTECT, ARROWS, STACHEM,
3D-UNDERWORLD, NOPTILUS, CURE) are focusing on the development of tools
and techniques to survey, assess, stabilize, monitor and preserve underwater archaeo‐
logical sites using robot systems and scanners. It is important to note that the majority
of these projects are not dealing with dissemination and wider public awareness of
underwater cultural heritage. In particular, in the Mediterranean Sea where most of the
oldest underwater assets exist, are at most in risk due to the marine environment,
trawlers, looting and wood degrading marine borers. Hence recording and promoting
CH is most important in the Mediterranean Sea than any other place.
1.1 Scope
Project’s iMARECULTURE scope is to raise public awareness of European identity by
focusing in maritime cultural heritage, which by default brings together different civi‐
lizations. In particular, it aims in bringing inherently unreachable underwater cultural
heritage, within digital reach of the wide public, by implementing virtual visits, serious
games with immersive technologies and underwater augmented reality. Scope of the
project is to design, analyze, develop and validate pioneer applications and systems in
the context of Virtual Museums through collaborative and innovative research from a
diverse group of scientists, researchers, archaeologists, experts and museums.
The project combines the two aforementioned groups of research projects into one
by merging advancements in VR with the underwater environment. It will accomplish
it by using existing technology to create breakthrough applications and digital experi‐
ences in the area of Virtual Museums in order to empower different types of users to
engage with European underwater cultural heritage digital resources by exploiting re-
use and re-purposing of existing data.
1.2 Approach
Submitted in the call ‘Virtual museums and social platform on European digital heritage,
memory, identity and cultural interaction’ (CULT-COOP-08-2016), the projects inves‐
tigate new ways to personalize the museum visit to a digital of physical visitor, while
Project iMARECULTURE: Advanced VR, iMmersive Serious Games 807
support social cohesion and European identity. Virtual museums are particularly strong
in visualizing CH that it either intangible, does not exist anymore, it is partially destroyed
or it is remotely located. Ancient maritime commerce is a perfect example of civiliza‐
tions’ interaction and cultural exchange, but unfortunately not easily exhibited to the
wider public. Ships, shipwreck sites and underwater sites in general, are far from public’s
reach and understanding. Enabling immersive technologies to allow for content
enhanced dry visits of visitors on such sites, it will inevitably raise public’s awareness
and stir further interest about maritime culture.
The goal is to bring shipwreck sites to the reach of the wider public, so that they can
have a personalized and interactive dry visit using VR googles from the comfort of their
house. Museum visits could be further enhanced, using immersive technologies, such
as VR caves and holographic screens. The latter allow for all visitors to witness an
interactive 3D experience of another visitor, while they carry no special glasses or any
other equipment, while freely walking around the objects under investigation. Moreover,
the project will enhance the underwater experience of diver visitors, as there are
submerged archaeological sites, that support such visits. Especially designed underwater
tablets using Augmented Reality (AR) will superimpose information about specific finds
and architectural designs on the screen of a specially designed underwater tablet. The
information will be provided in real time, and on user’s demand, in order to avoid over‐
loading the diver.
Following the need to extend any visit, pre- and after- the visit serious games and
storytelling, encourage and surrounds the physical visit. As a pre-visit experience, social
platform users will be able to participate in a seafaring game, sailing between ancient
ports for commerce, trying to confront all natural and human hazards the ancient crew
had to face, using the limited resources and limitations of that specific time period.
Similarly, as an after-visit experience, the social platform users will be able to partner
up for a virtual excavation dive on an ancient shipwreck, with all limitations and prob‐
lems that such dives have, and enjoy the ‘discovery’ and surfacing of new artifacts, while
appreciate the work done by underwater archaeologists.
Three sites have been carefully selected for project’s implementation, based on their
ability to support the context of each action, as well as their data availability, so that no
assets were to be allocated on data acquisition.
Mazotos shipwreck site: The site [1], when discovered, was virtually undisturbed,
so its archaeological importance, as well as the immediate need for its protection, trig‐
gered the organization of the first Cypriot underwater archaeological project. The wreck
lies at a depth of −44 m, some 14 nautical miles south-west of Larnaca, 1.5 nm from
the shore, near Mazotos village. The main visible feature of the site is a concentration
of amphorae on a sandy, almost flat sea-bed. Its maximum vertical relief measures 1 m
and its maximum dimensions are 17.5 × 8 m. The oblong concentration, almost in the
form of a ship, has a north-south orientation and consists of 500–800 Chian amphorae
partly or totally visible, which date to the middle of fourth century BC.
Xlendi shipwreck site: The Phoenician shipwreck off Gozo Xlendi. resting at a depth
of 110 m, it is probably the oldest ancient shipwreck in the central Mediterranean
(700BC). Xlendi Bay is a narrow inlet on the south-west coast of Gozo, close to the
island’s south-western point Rasil-Wardija and to the capital Rabat, which is 2.5 km away.
Rabat is a hilltop town that is thought to have been the main settlement area throughout
Gozo’s history. Xlendi lies at the mouth of a valley which runs from the Rabat hill to the
coast. The valley is bounded by high rock walls but opens into a floodplain, which is the
site of the modern village, before meeting the sea. There is a good, fresh water source in
the village of Fontana, on the road between Rabat and Xlendi [2, 3].
Baiae underwater archaeological park: The Baiae site is of particular interest, it
was created in 2002 and it is together a Marine Protected Area and an Underwater
Archaeological Park. Environmental aspects of this area are related to a peculiar volcanic
and deformational history. The submerged area includes part of the territory of the
ancient city of Baiae and Portus Iulius, comprising the roman harbour and numerous
constructions used as warehouses. The archaeological remains include luxurious mari‐
time villas and imperial buildings, more modest houses, private thermae, tabernae and
all those structures that characterize the cities of the Roman age. The itineraries of “Villa
con ingresso a protiro” (selected for this Project) with black and and white and white
mosaic floors, and thermae; the Villa of Pisoni, the Nymphaeum of Punta Epitaffio (with
copies of statues of the Imperial families and of Roman gods); the Portus Iulius with the
remains of mosaics floor of a republican Villa and a building with a porticos. The Baiae
underwater archaeological park will serve also as a demonstration site for the developed
Augmented Reality systems.
All actions will be evaluated in order to attain measurable results, and enhanced by
appropriate storytelling.
2 Project Approach and Methodology
2.1 Objectives
Being an interdisciplinary project, the first step is the clear definition of targets, goals,
test sites and wrecks to be used. Several decisions about the serious games, the age, the
era and data to be gathered, will be taken. Discussions about existing VR and AR tech‐
nologies [4] and implementation will clarify the roadmap towards milestones and goals.
Data gathering and pre-processing phase: Is the fundamental pillar for the project.
Both 3D as well as supporting data will be accumulated by partners, from open sites,
published journals and books and partner’s archives to support storytelling, narratives,
3D models of sites [5–9], ships [10], cargos, probabilistic geospatial analysis about ship
routes [11], wreck site formation processes, etc. Having all the necessary information,
iMARECULTURE will create and re-use a plethora of 3D models for both the ship
wreckages and the amphorae, allowing people to examine the 3D underwater environ‐
ment and objects on it. Similarly, a site of Baiae underwater archaeological park where
a complete 3D model exists [12, 13], will be used for the implementation of the AR
supported dives.
Virtual reality phase: This phase offers the users more than underwater tour, at a safe
environment. Additionally, this phase offers the enhanced experience of the diver, visitor
or scholar, using underwater augmented reality in a tablet and an underwater housing.
In VR people around the world can access any environments using immersive
technologies and internet access. Supportive narrative and multimedia tools [14] will
provide interactive information about the site and objects allowing for a personalized
experience, both in terms of navigation as well as in objects’ queries. At the same time,
emphasis will be given towards the creation of a universal standard for storytelling
presentation [15]. End-users will be in position to interfere/adjust the excavation
discovery story, according to their liking.
Serious games in platforms: All the knowledge acquired through this project will lead
to the creation of serious games in platforms [16, 17]; actual and high detailed environ‐
ments will be presented, allowing the gamers to fully explore and interact with them.
The main focus will be on the two imperative aspects of the serious games: storytelling
and interaction; both of which are essential in engaging the users and inspiring further
intrinsic learning.
Additionally, the knowledge base will cover, under a unified framework, a variety
of underwater CH sites of different properties and characteristics. Holistic records enable
rich and generic characterization of the aforementioned cultural heritage elements into
well-defined and structured components and machine-readable formats that allow the
utilization of multi-domain information in an automated way. This holistic knowledge
is built upon the CIDOC/CRM protocol and it extends the standard to include historic
documentation, geometric survey, material survey, as well as critical environmental
parameters. Furthermore, the project will extent BIM protocol, which is a process
involving the generation and management of digital representations of physical and
functional characteristics of places and objects, in order to describe and manage knowl‐
edge about underwater sites and CH assets.
Finally, diagnostic schemes for site formation analysis will be bullied by incorpo‐
rating (i) underwater changes profiling, (ii) non-invasive/non-destructive analysis
methods while (iii) documenting the sensor acquired spatio-temporal data and (iv)
supporting data aggregation strategies. Protocols developed in past EU projects will
form the basis for the project protocols and will be adapted according to specifications.
3 Methodology
3.1 IMARECULTURE Methodology Steps
Figure 1 is presenting the methodology steps adopted for this project. The project starts
with the project scenario and use case definition. The cultural heritage items of the three
involved underwater cultural sites will be selected in a way that covers all iMARE‐
CULTURE challenges. Then, an iterative process is adopted to specify a concrete archi‐
tecture that can fulfil all user requirements and pilot sites demands. The design is
performed at two cycles. The first gives the first form of the algorithms needed to be
applied.
Fig. 1. The i-MareCulture methodology steps
Additionally, are specified and defined the IPR issues involved in the project. Then
the first set of algorithms is launched and the updated architecture takes into account
results from the previews step. Subsequently the second set of algorithms is launched.
At the next step the first integrated and validated is created: All the aforementioned
research components are integrated to launch the first initial prototype.
Then we launch the final set of algorithms and we design the pilots and the evaluation
frameworks used for objectively assess the performance and acceptance of the archi‐
tecture. Finally, the final iMARECULTURE platform will be assessed by different types
of users.
3.2 Description of “Virtual Sites”
The project exploits 3 different “Virtual Sites” (Fig. 2) (existing sites for virtual and
augmented reality applications) in order to implement its multidisciplinary approach
and accomplish the described objectives. The proposed sites are representative of
different kind of Underwater Cultural Heritage, of different states of environmental and
geomorphologic conditions (i.e. water depth) and of different periods, in order to present
the users a wide range of the common European maritime culture. Consequently, our
selection of 3 virtual sites includes: (1) Mazotos shipwreck site, Cyprus; (2) Xlendi
shipwreck site, Malta and (3) Baiae underwater archaeological park, Italy. These cover
the Mediterranean basin from center to east. We have also taken care that these sites
cover a wide chronological span from 700BC to 400AD.
(a) (b) (c)
Fig. 2. (a) The main concentration of Mazotos shipwreck site (photo: MARELab, University of
Cyprus), (b) Amphorae laying on the seabed at the Xlendi wreck (Photo: The University of Malta/
COMEX/CNRS, (c) The Nymphaeum of Punta Epitaffio
3.3 Description of Demonstration Sites
Thalassa Museum: Thalassa Municipal Museum was opened in August 2005, and it is
directed by the Pierides Foundation, in association with the Hellenic Institute for the
Preservation of Nautical Tradition and the Tornaritis-Pierides Marine Life Foundation.
The Museum is located at the centre of Agia Napa in Cyprus and it has the sea as a
subject-matter. It is the first museum of its kind across the Mediterranean region with
main principle to present to the audiences, the local and foreign visitors, the impact and
the significance of the sea upon the history of the island.
The main exhibit at the fourth level and of the museum itself is the ‘Kyrenia II’ vessel
(Fig. 3). A life size exact replica of the ancient ship of Kyrenia of the Classical period
(400 B.C.), which was built in 1985 for scientific experimental purposes by the Hellenic
Institute for the Preservation of Nautical Tradition of Athens. At the same level the
visitor walks on a glass floor where a reconstruction of the old shipwreck is displayed.
Fig. 3. The life size exact replica of the ancient ship of Kyrenia, the reconstruction of the old
shipwreck
4 Specific Innovations
The project connects education, research and industry by supporting and boosting inno‐
vative enterprise to develop their technological breakthroughs into viable products in
the area of Virtual Museums and Digital Heritage, with real commercial potential.
Furthermore, based on the scope and specific aims of the project, significant research
will be conducted in
• Developing the outline and short population of a 3D Library as tools for Maritime
Archaeology using ontological schemes
• Digitization of naval engineering models and shipbuilding content
• Geospatial analysis of naval routes among main ancient commerce centers
• Multimedia storytelling and interaction
• Underwater Augmented Reality Interfaces
• Hybrid underwater tracking solution based on acoustic sensors and computer vision
• Virtual Reality Interfaces on VR glasses, VR caves and holographic screens
• Serious games through social platforms
The tools will be validated and tested across real-world application scenarios and
cases and under a number of different participants. This way, we will prove the inno‐
vation potential of our platform over diverse and challenging environments. The serious
games platform, the AR/VR interfaces will be assessed in real cases and will include
scenarios coming from maritime archaeology research so as to achieve a great mixture
between entertainment, informal educational, and underwater/maritime CH under‐
standing. The project supports concrete plans and exploitation activities at certain times
to further improve innovation potential.
5 Conclusions
The project iMARECULTURE combines different research disciplines, namely experts

in 3D acquisition, Virtual and Augmented reality, serious games developers, geo-statis‐
tics and GIS, archaeologists, story tellers, along with technology of underwater tablets,
holographic screens, in underwater sites and museums into one group. This way it will
accomplish existing technology to create breakthrough applications and digital experi‐
ences in the area of Virtual Museums in order to empower different types of users to
engage with European underwater cultural heritage digital resources by exploiting re-
use and re-purposing of existing data.
References
1. Demesticha, S.: The 4th-Century-BC Mazotos Shipwreck, Cyprus: a preliminary report. Intl.
J. Naut. Archaeol. 40(1), 39–59 (2011)
2. Drap, P., Merad, D., Hijazi, B., Gaoua, L., Nawaf, M.M., Saccone, M., Chemisky, B.,
Seinturier, J., Sourisseau, J.-C., Gambin, T.: Underwater photogrammetry and object
modeling: a case study of Xlendi Wreck in Malta. Sensors-Basel 15, 30351–30384 (2015)
3. Azzopardi, E.: The shipwrecks of Xlendi Bay, Gozo, Malta. Intl. J. Naut. Archaeol. 42, 286–
295 (2013). doi:10.1111/1095-9270.12020
4. Liarokapis, F.: An augmented reality interface for visualizing and interacting with virtual
content. Virtual Reality 11(1), 23–43 (2007). Springer
5. Poullis, C.: A framework for automatic modeling from pointcloud data. IEEE Trans. Pattern
Anal. Mach. Intell. 35(11), 2563–2575 (2013)
6. Demesticha, S., Skarlatos, D., Neophytou, A.: The 4th-century BC shipwreck at Mazotos,
Cyprus: new techniques and methodologies in the 3D mapping of shipwreck excavations. J.
Field Archaeol. 39(2), 134–150 (2014)
7. Georgopoulos, A., Agrafiotis, P.: Documentation of a submerged monument using improved
two media techniques. In: 2012 18th International Conference on Virtual Systems and
Multimedia (VSMM), pp. 173–180. IEEE (2012)
8. Balletti, C., Beltrame, C., Costa, E., Guerra, F., Vernier, P.: 3D reconstruction of marble
shipwreck cargoes based on underwater multi-image photogrammetry. Digit. Appl. Archaeol.
Cult. Heritage 3(1), 1–8 (2016)
9. McCarthy, J., Benjamin, J.: Multi-image photogrammetry for underwater archaeological site
recording: an accessible, diver-based approach. J. Marit. Archaeol. 9(1), 95–114 (2014)
10. Castro, F., Gomes-Dias, D.: Moulds, Graminhos and Ribbands: a pilot study of the
construction of saveiros in Valença and the Baía de Todos os Santos area. Brazil. Intl. J. Naut.
Archaeol. 44(2), 410–422 (2015)
11. Leidwanger, J.: Modeling distance with time in ancient Mediterranean seafaring: a GIS
application for the interpretation of maritime connectivity. J. Archaeol. Sci. 40, 3302–3308
(2013)
12. Bruno, F., Lagudi, A., Gallo, A., Muzzupappa, M., Davidde, B.P., Passaro, S.: 3D
documentation of archaeological remains in the underwater Park of Baiae. Intl. Arch.
Photogramm. Remote Sens. Spatial Inf. Sci. - ISPRS Arch. 40(5W5), 41–46 (2015)
13. Petriaggi, B.D., de Ayala, G.G.: Laser scanner reliefs of selected archaeological structures in
the submerged Baiae (Naples). Intl. Arch. Photogramm. Remote Sens. Spatial Inf. Sci. 5, 79–
83 (2015). doi:10.5194/isprsarchives-XL-5-W5-79-2015
14. Balogh, T., Forgacs, T., Agocs, T., Bouvier, E., Bettio, F., Gobbetti, E., Zanetti, G.: A large
scale interactive holographic display. In: IEEE VR2006 Conference, Virginia, USA (2006)
15. Rizvic, S., Skalonjic, I.: Reconstructing cultural heritage objects from storytelling - virtual
presentation of Sultan Murat’s Fortress in Sjenica. In: Proceedings of 2nd International
Congress on Digital Heritage, vol. 2, ISBN:978-1-5090-0047-0 (2015)
16. Anderson, E.F., McLoughlin, L., Liarokapis, F., Peters, C., Petridis, P., de Freitas, S.:
Developing serious games for cultural heritage: a state-of-the-art review. Virtual Reality
14(4), 255–275 (2010). Springer
17. Mortara, M., Catalano, C.E., Bellotti, F., Fiucci, G., Houry-Panchetti, M., Petridis, P.:
Learning cultural heritage by serious games. J. Cult. Heritage 15(3), 318–325 (2014)
Immersivity and Playability Evaluation
of a Game Experience in Cultural Heritage
Roberto Andreoli, Angela Corolla, Armando Faggiano, Delfina Malandrino(B) ,

Donato Pirozzi, Mirta Ranaldi, Gianluca Santangelo, and Vittorio Scarano
Department of Computer Science, University of Salerno,

Via Giovanni Paolo II, 132, 84084 Fisciano, SA, Italy
{dmalandrino,dpirozzi,vitsca}@unisa.it
Abstract. The introduction in the market of head-mounted displays

(HDMs), originally used for gaming, opens the door to a wide set of
application fields that could benefit of characteristics, such as immersiv-
ity, presence as well as a high degree of realism. In the field of Cultural
Heritage, an immersive virtual experience can enhance playfulness and
involvement in the fruition of a cultural experience, by determining a
more efficient knowledge absorption and retention of the learnt content.
In this work we introduce a prototype of a Serious Game in Cultural
Heritage, named HippocraticaCivitasGame, designed and implemented
to foster playfulness and learning effectiveness. We also performed an
evaluation study to assess users’ perceived immersivity and playability,
as well as the effectiveness when analyzing the acquired knowledge about
the archaeological site structure and the proposed learning goal.
Keywords: Virtual reality in cultural heritage · Serious game

technologies · Evaluation
1 Introduction
Immersive Virtual Reality (VR) has been widely used in several domains ranging
from entertainment and gaming to education and training (military, aerospace
and defense, health, and so on), from heritage and archeology to architectural
design, from therapy to retail and interactive media [1,3,8].
According to a new forecast from market research firm Gartner [10], wearable
electronic device sales will increase 18.4 % year over year to hit 322.7 million
shipments in 2017 ($6.31 million coming from head-mounted displays, with a
percentage increase of 341 % from 2016 to 2017). The demand for virtual reality
devices will grow rapidly in the coming years owing to the wide availability of
technologically advanced virtual reality devices, their increased computational
power and reduced costs, and their increasing utility in a variety of applications.
Indeed, some of the big technology giants have entered the virtual reality mar-
ket. Facebook, Inc. acquired Oculus VR, in March 2014 for approximately $2

DOI: 10.1007/978-3-319-48496-9 65
Immersive Experiences in Cultural Heritage 815
billion. Sony, HTC and Microsoft Corporations released their virtual displays,
that is the PlayStation VR headset, the HTC Vive and the Microsoft HoloLens
devices, respectively. The aim for all is the same: “to further extend virtual real-
ity into verticals beyond gaming, such as communications, media, entertainment,
education and other areas”.
Moreover, the 2016 Horizon Report from the New Media Consortium [12] pre-
dicted that immersive virtual reality technology is likely to have a large impact
on teaching and learning in higher education worldwide in the next years. The
compelling aspect of virtual reality is its ability to transport learners into envi-
ronments and situations that they otherwise may never have access to, being
able to be in the present, but projected in the past. Serious Games (SGs) [23],
which are focused on the learning aspect of games, combined with immersive
technologies are able to improve players’ experience, therefore contributing to a
more effective knowledge acquisition and retention [7].
The objective of this work is to present HippocraticaCivitasGame, a proto-
type of a Serious Game in Cultural Heritage, designed and developed to fos-
ter the understanding and the learning of an archaeological site, set at Palazzo
Fruscione-San Pietro a Corte in Salerno, and of its historical elements in effective
and enjoyable way. We also performed an evaluation study to assess effectiveness
in terms of knowledge acquisition of a specific learning goal, the game playability,
and finally users’ appreciation and acceptance of the developed SG.
2 Related Work
Several relevant works are available in literature, including works addressing

immersivity and effectiveness of knowledge acquisition and retention. Immer-
sion, in its several acceptations, has been studied in different works [18,20].
Moreover, some other works addressed other interesting game characteristics,
such as, perceived presence and usability [11,14].
Park et al. [21] designed a system mediated by HMD technology to experience
an immersive tour in Cultural Heritage. Immersiveness was positively rated by
players when using virtual devices compared to the traditional tour method.
About the correlation among immersivity, emotions and knowledge retention,
Chittaro et al. [7] considered the use of a visor for a SG in aviation safety, allowing
players experiencing an emergency evacuation on an aircraft compared with the
traditional safety card. Authors observed that the immersive game increases the
knowledge retention as well as the user engagement.
Bellotti et al. [2] proposed a model abstracting the SandBox SG, which offers
a generalization of task-based learning theory. The authors developed a SG in
CH field, named Travel in Europe, assessing the knowledge acquisition and con-
cluding that SGs represent a significant opportunity for learning.
A well-known and crucial limit to mention is the cybersickness effect caused
by virtual reality devices during their usage. Specifically, cybersickness is caused
by a disagreement between visually perceived movement and the vestibular sys-
tem’s sense of movement.
816 R. Andreoli et al.
Various studies indicate it can affect 50 % to 80 % of people, depending on

the fidelity of the digital content and how it is presented [6]. Works in this
field state that the problem can often be improved with habituation [16] or by
intervening on the field of view of used displays [5]. Lee et al. [17] investigated the
cross-dimensional virtual reality coupling the Leap Motion for the hand tracking
with the Oculus Rift. As result, the mixed technology with the seated position,
significantly reduced nausea.
3 A Serious Game for CH

In this section we present the prototype of our immersive Serious Game, named
HippocraticaCivitasGame(a screenshot is shown in Fig. 1). First, we introduce
the game design, and then we discuss about some technical details of the game
implementation.
3.1 Game Design

In the old town of Salerno, there is the historical structure of S. Pietro a Corte-
Palazzo Fruscione. It is an archaeological site that shows how the city developed
since Roman times until nowadays. The most important elements are the remains
of the palace of the Longobard Prince Arechis II, which constitutes a rare pre-
served example of civil architecture of the eighth century A.D. [22]. The site was
investigated through excavations from 1988 to 1990, and recently in 2011 and
2012. The archaeologists unearthed the remains of Roman buildings interpreted
as Thermae of I-II century A.D. to above which was erected the Arechis’s palace
with a richly decorated chapel. In the Norman period, just to the north, was
built another mansion, which was expanded under the Angevin kings and then
it was transformed to the present day.
Fig. 1. HippocraticaCivitasGame snapshot of the male frigidarium.
In order to design the game, we identified the game macro learning goal,
which is the understanding of archaeological site pluri-stratification. Indeed, it
Fig. 2. Thermae level map. Atrium, vestibule and room with cabinet (0-1-2); female
apodyterium (F1), frigidarium (F2), tepidarium (F3), caldarium (F4); male apody-
terium (M1), frigidarium (M2), tepidarium (M3), caldarium (M4); praefurnium (P).
has different historical periods, the Roman, Longobard, Norman and Angevin
periods, with different architectural styles as well as a variety of archaeological
and historical elements. When playing, users through the posed challenges can
experience, visit and hopefully grasp such a variety. In particular, the game is
divided in three levels, one for each historical period. Micro learning goals are
single challenges to solve during the game to learn specific environment, historical
or archaeological elements. All the micro learning goals together contribute to
the whole archaeological site understanding.
In order to evaluate game immersivity, we focused on the first level which
corresponds to the Roman Thermae. Therefore, the game starts in the vestibule
of the the Virtual Thermae environment (Fig. 2, room 1). The environment,
is divided into female and male parts; each part is composed of four rooms:
the changing room (apodyterium), the cold room (frigidarium), temperate room
(tepidarium) and hot room (calidarium). This level, firstly, aims to show how
thermal environments were structured. Thus, as narrative pretext we introduced
the challenge to bring an object from one place to another. This challenge, a well-
known game design pattern [4], can be exploited to foster the visit of the places
in a map. The object to find was the strigil (i.e., a tool for body cleansing) placed
within a cabinet (Fig. 2, room 2) and to move in one of the two apodyterium.
3.2 Game Implementation
The HippocraticaCivitasGame is based on Unity, a cross-platform game engine

that offers many features and extensions to develop games. In addition to the
game engine features, we designed a collection of reusable building blocks to facil-
itate the future extensions. The main building blocks are: Controller Manager,
Game Logic, Game State, Storage, and Display Manager.
Players interact with our Serious Game using the Oculus Rift (DK2) and the
gamepad XBOX 360 controller devices. Inputs from those devices are forwarded
to the Controller Manager, which is responsible for their processing, and for
generating a command directed to the Game Logic. For instance, when the player
presses a specific button on the gamepad, this input is associated to an avatar
action; the same action can be triggered by a different input device, such as
the Microsoft Kinect by envisioning a specific gesture. Therefore, the Controller
Manager is an interface between the input devices and the Game Logic. Of
course, the action may change the world and the avatar state (e.g., rotating it).
The Game Logic coordinates the different game components and defines the
game rules. It handles the starting of a new game, avatar initial position, and
initialisation of game objects (e.g., the strigil). It contains a timer to keep track
of the game duration, that lasts for five minutes; when time expired the avatar
is blocked and the data structure that contains avatar’s statistics (e.g., move-
ments on the map) are saved in a file; for performance reasons, this operation is
performed, only once, at the end of the game.
The Game State contains the entire game status in terms of avatar state and
world state. It stores, every second, the avatar position, rotation and time in the
Storage; it also keeps track of objects’ gathering and positioning, e.g., the strigil
object in our Serious Game. Finally, the main goal of the Display Manager is to
render the game; it shows textual and image information during the immersive
experience. Finally, a demo of HyppocraticaCivitasGame, with a three-minutes
visit in Palazzo Fruscione-San Pietro a Corte, is available online1 .
4 Evaluation
In this section we first describe the methodology that we employed for our evalu-
ation study, defined according to the standard HCI methodology [13], afterwards
we discuss the results obtained when a group of 72 people was involved in testing
our Serious Game.
4.1 Methodology
Our study took place in a large, dedicated space in a research laboratory at the
University of Salerno. The workstation used for rendering the environment was
equipped with an i7-4770K 8-CPU, a Nvidia GTX770 Graphics card, and 16 GB
of main memory. The input device used was a wireless Microsoft XBOX 360
Controller. The virtual reality headset supported was based on the Oculus Rift.
The study envisioned three different phases in which we carried out: (a) a
Preliminary Survey (b) a Testing Phase and (c) a Summary Survey, as defined
and implemented in other contexts [9,19].
As first step, participants filled out a preliminary survey questionnaire about
their demographics (i.e., gender, age, education level), video gaming experience,
and eventual sensitivity to the kinetosis or motion sickness problem, that is the
disagreement that exists between visually perceived movement and the vestibular
system’s sense of movement. Thereafter, participants were briefly informed about
1
http://www.di.unisa.it/∼delmal/research/usability/Euromed2016/Demo/.
the task to perform. Specifically, the challenge for the players is to find the strigil
in the game environment and place it in one of the two apodyterium, as described
in Sect. 3.1. They had a fixed amount of time to complete the task (i.e., 5 min).
When this time expired, interested participants were anyway allowed to continue
the virtual experience. At the end of the testing phase we asked users to fill in a
questionnaire, adapted from [24], aiming at assess the perceived degree of pres-
ence, control and motivation, and playability characteristics. Performance data
were collected during the gaming experience. Specifically, we logged, for each
participant, the time when she started the experience, the time when she found
the strigil, the time when it was arranged in the requested place, and finally, the
time when the experience finished. In the last phase we asked users to fill in a
summary survey aiming at assess the effectiveness of the knowledge acquisition
with regard to the performed task. We also gathered users’ perceptions about
the usefulness of the serious game for educational purposes and their propensity
to recommend the game to a friend.
To analyze the opinions of participants in terms of immersion and their overall
subjective experience we used the standard Presence Questionnaire (PQ) [24].
It measures in what degree users feel as they are actually part of the experi-
enced environment. From this questionnaire we extracted questions that allow
to obtain sub-scores in terms of three factors: realism (RF), control factor (CF),
and involvement (INV). We also studied cybersickness, the digital version of
the motion sickness, by administering a questionnaire adapted from [15]. All
questionnaires, translated into Italian, are available online2 , while the whole
evaluation study lasted approximately 40 min.
Subjects were students, teachers, and employers at several departments of
the University of Salerno. They were recruited through flyers and word of mouth
advertising. Their participation was voluntary and anonymous. They were also
informed that they could stop at any time during the evaluation process. For
data analysis, non-parametric tests (Kruskal-Wallis and Mann-Whitney) were
applied to study differences between groups. The Shapiro-Wilk goodness-of-fit
test. Finally, questionnaire responses were analyzed using SPSS.
4.2 Results
A total of 72 participants were recruited from the general campus population

(49 % Human field study, 51 % Scientific field study). In detail, the education
level included 25 % with bachelor degree, 69 % with a master degree, and 6 %
with a PhD. The sample was mostly male (64 % male, 36 % female) with a
mean age of 27 (SD = 8). 67 % of participants reported intermediate or higher
experience with videogames, 21 % stated to be expert, while only 12 % reported
a low experience in that field. 35 % reported to be prone to motion sickness (we
did not find any statistical difference with regard to the gender factor). Finally,
the entire sample had no confidence with the Oculus Rift device.
2
http://www.di.unisa.it/∼delmal/research/usability/Euromed2016/Questionnaires.
pdf.
Results of the Testing phase showed that 80 % of participants was able to find
the strigil, but only 64 % was able to put it in the right place. On average, it took
about 2.5 min to find the strigil, and about 63 seconds to put back it in the right
place. However, almost all participants (96 %) were able to successfully respond
to the question about the definition of the strigil (among three different provided
answers). This result suggests that participants paid attention to the messages
conveyed during the experience, although they stated to be very involved in the
game. The overall result indicated a positive learning outcome for the knowledge
question posed.
Results about the SG quality characteristics are shown in Fig. 3(a) and
Table 1. Specifically, in Fig. 3(a) we report the results of users’ perceptions about
playability, the control they had within the game, and their motivation. Partici-
pants enjoyed the experience (M = 6.0, SD = 1.0). They also rated positively the
aesthetics of the Serious Game (M = 5.9, SD = 1.3). We did not find any statisti-
cal differences with regard to these metrics when considering the gender factor.
About the Control factor, our sample agreed with how easy was to use the Ocu-
lus (“I found the field of view rotation using Oculus Rift - head movement - very
easy to perform”, M = 6.1, SD = 1.4) and expressed high satisfaction in terms of
the sensitivity of the controller (“The game controller sensitivity in the virtual
experience was adequate”, M = 6.0, SD = 1.1). Similarly, the Motivation factor
was positively rated by our sample (M = 6.5, SD = 0.7).
As shown in Table 1, the Presence questionnaire is aggregated in four fac-
tors, all positively rated. Participants found out responsive the overall system
(M = 6.5, SD = 0.8), and they perceived very natural the interactions within the
game (M = 6.0, SD = 1.0). What is really remarkable is that, despite participants
had few time to play, they rated very positively their involvement as well as their
increased ability at the end of the experience (M = 5.6, SD = 1.5), highlighting
the easiness of the controls that we designed for the game. The Realism Factor
was highly positively rated, as also derived by informal suggestions at the end of
the evaluation process. 35 % of participants, in fact, provided as positive aspect
of the overall experience, the quality of the rendering and how much “real” the
virtual experience was perceived. Finally, we did not find any statistical differ-
ences with regard to these metrics when considering the gender factor. Instead,
we found out a significant correlation among all the analyzed subscales.
Results about cybersickness, showed that dizziness was the symptom most
felt by the participants at the study (44 %). Nausea was the second most felt
symptom, with the highest degree of severity (7 % felt the nausea with a sever
level). Only 20 % of participants did not suffered of any type of problem. More-
over, when comparing kinetosis results (question posed in the Preliminary sur-
vey) with cybersickness results (question posed in the Summary Survey) we
found out that 43 %, out of 65 % who reported do not suffer of any motion sick-
ness problem, experienced cybersickness during the immersive experience. We
did not find differences when analyzing cybersickness in terms of demographic
factor.
As further analysis, we classified participants in three groups, namely,
“Young” (with the age range 19–25), “Young Adults” (with the age range 26–31),
Table 1. Presence and Immersion. 7-Point Likert scale.
ID Question Factor Mean SD

1 How responsive was the environment to actions CF 6.5 0.8
that you initiated (or performed?)
2 How natural did your interactions with the envi- CF 6.0 1.0
ronment seem?
3 How much did your experiences in the virtual RF 5.6 1.1
environment seem consistent with your real-world
experiences?
4 How involved were you in the virtual environment INV 6.3 1.0
experience?
5 How proficient in moving and interacting with the CF 5.6 1.5
virtual environment did you feel at the end of the
experience?
6 Were you involved in the experimental task to the INV 5.6 0.5
extent that you lost track of time?
and “Adults” (with the age range 31–60). Our Adults users were also the less
familiar with video games and immersive systems (4 experts in the older group
against 11 in the young one). Interestingly, we found out a correlation between
age and reported sense of general discomforts due to cybersickness. We also
found out a correlation between the feeling of discomforts due to cybersickness
and the experience of the users with virtual immersive environments. Oculus Rift
devices have not completely penetrated the consumer market, but as discussed
in Sect. 1, they will become more and more popular in the next few years. As a
consequence, users will start to get used to them, and therefore, cybersickness
Fig. 3. Results of the presence and simulator sickness questionnaires.

could be reduced through training and established experience. However, a lon-

gitudinal study, with a large and diversified sample of participants, is needed to
study these effects. Finally, we did not find any statistical differences among the
three groups of users in terms of playability, control and motivation factors.
Results of the Summary survey showed that the use of games for Cultural
Heritage was highly accepted by participants. Specifically, on average, both the
usefulness of the idea of using the SG for learning (M = 6.5, SD = 0.8), and
the propensity to recommend to other the experience (M = 6.7, SD = 0.6) were
positively rated.

This paper proposed HippocraticaCivitasGame, a Serious Game prototype
designed and developed to foster learning and understanding of the Roman Ther-
mae in the Palazzo Fruscione-San Pietro a Corte, in Salerno. The performed
evaluation study showed that our SG was able to convey important information
about the structure of the site and the micro learning goal analyzed. Indeed,
96 % of participants successfully responded to the question about the definition
of the strigil, among the three different provided answers.
Results about presence and immersion revealed that participants enjoyed
the experience and they also rated positively the aesthetics of the whole Serious
Game. 35 % of participants provided as positive aspect of the overall experience,
the quality of the rendered environment, and how much real the virtual expe-
rience was perceived. Moreover, despite they had few time to play, they rated
very positively their involvement as well as their increased ability at the end of
the experience, highlighting the easiness of the controls that we designed for the
game. Furthermore, the idea of using our SG for learning was highly accepted
by participants. Specifically, on average, both usefulness and the propensity to
recommend to other the experience were positively rated.
Finally, we are currently working on how to extend the SG with other game
levels and with the corresponding micro learning objectives. Additionally, we
are enhancing the SG to allow collaboration among multiple players, in order to
further guarantee involvement and playability. An extensive, representative, and
longitudinal evaluation study of the final game will be performed accordingly.
Acknowledgments. This research was supported by the DATABENC (Distretto ad

Alta Tecnologia per i Beni Culturali) project, REA NA-876990, Regione Campania,
Italy. We would also like to thank all participants that took part at the study.
References
1. Adding a level of reality to online shopping. http://www.starthub.org/news/
adding-level-reality-online-shopping. Accessed 10 Jun 2016
2. Bellotti, F., Berta, R., De Gloria, A., D’ursi, A., Fiore, V.: A serious game model
for cultural heritage. J. Comput. Cult. Heritage (JOCCH) 5(4), 17 (2012)
3. Bellotti, F., Berta, R., Gloria, A.D., Panizza, G., Primavera, L.: Designing cultural
heritage contents for serious virtual worlds. In: International Conference on Virtual
Systems and Multimedia, pp. 227–231 (2009)
4. Bjork, S., Holopainen, J.: Patterns in game design (2004)
5. Carnegie, K., Rhee, T.: Reducing visual discomfort with hmds using dynamic depth
of field. IEEE Comput. Graph. Appl. 35(5), 34–41 (2015)
6. Chang, C.H., Pan, W.W., Tseng, L.Y., Stoffregen, T.A.: Postural activity and
motion sickness during video game play in children and adults. Exp. Brain Res.
217(2), 299–309 (2012)
7. Chittaro, L., Buttussi, F.: Assessing knowledge retention of an immersive seri-
ous game vs. a traditional education method in aviation safety. IEEE Trans. Vis.
Comput. Graph. 21(4), 529–538 (2015)
8. Cukor, J., Gerardi, M., Alley, S., Reist, C., Roy, M., Rothbaum, O.B., Difede, J.,
Rizzo, A.: Virtual reality exposure therapy for combat-related PTSD. In: Ritchie,
E.C. (ed.) Posttraumatic Stress Disorder and Related Diseases in Combat Veterans,
pp. 69–83. Springer International Publishing, Switzerland (2015)
9. Fish, A., Gargiulo, C., Malandrino, D., Pirozzi, D., Scarano, V.: Visual exploration
system in an industrial context. IEEE Trans. Ind. Inf. 12(2), 567–575 (2016)
10. Gartner: Worldwide Wearable Devices Sales Forecast (2016). http://www.gartner.
com/newsroom/id/3198018. Accessed 13 Jun 2016
11. Hupont, I., Gracia, J., Sanagustı́n, L., Gracia, M.A.: How do new visual immersive
systems influence gaming QoE? a use case of serious gaming with Oculus Rift. In:
International Workshop on Quality of Multimedia Experience, pp. 1–6 (2015)
12. Johnson, L., Adams Becker, S., Cummins, M., Estrada, V., Freeman, A., Hall, C.:
NMC Horizon Report: 2016 Higher Education Edition (2016)
13. Jonathan, L., Feng, J.H., Hochheiser, H.: Research Methods in Human-Computer
Interaction. Wiley, Hoboken (2010)
14. Kalawsky, R.S.: The validity of presence as a reliable human performance metric
in immersive environments. In: proceedings of the Presence Workshop 2000 (2000)
15. Kennedy, R.S., Lane, N.E., Berbaum, K.S., Lilienthal, M.G.: Simulator sickness
questionnaire: an enhanced method for quantifying simulator sickness. Int. J.
Aviat. Psychol. 3(3), 203–220 (1993)
16. Kennedy, R.S., Stanney, K.M., Dunlap, W.P.: Duration and exposure to virtual
environments: sickness curves during and across sessions. Presence Teleoper. Vir-
tual Environ. 9(5), 463–472 (2000)
17. Lee, P.W., Wang, H.Y., Tung, Y.C., Lin, J.W., Valstar, A.: TranSection: hand-
based interaction for playing a game within a virtual reality game. In: Proceedings
of the 33rd Annual ACM Conference Extended Abstracts on Human Factors in
Computing Systems, CHI EA 2015, pp. 73–76 (2015)
18. Lugrin, J.L., Cavazza, M., Charles, F., Le Renard, M., Freeman, J., Lessiter, J.:
Immersive FPS games: user experience and performance. In: ImmersiveMe 2013,
pp. 7–12 (2013)
19. Malandrino, D., Scarano, V., Spinelli, R.: How increased awareness can impact atti-
tudes and behaviors toward online privacy protection. In: International Conference
on Social Computing, SocialCom 2013, pp. 57–62 (2013)
20. McMahan, R.P., Bowman, D.A., Zielinski, D.J., Brady, R.B.: Evaluating display
fidelity and interaction fidelity in a virtual reality game. IEEE Trans. Vis. Comput.
Graph. 18(4), 626–633 (2012)
21. Park, D., Nam, T.J., Shi, C.K.: Designing an immersive tour experience system
for cultural tour sites. In: CHI 2006 Extended Abstracts on Human Factors in
Computing Systems, pp. 1193–1198 (2006)
22. Peduto, P., Fiorillo, R., Corolla, A.: Salerno. Una sede ducale della Langobardia
meridionale. Spoleto: CISAM (2013)
23. Susi, T., Johannesson, M., Backlund, P.: Serious games: an overview (2007)
24. Witmer, B.G., Singer, M.J.: Measuring presence in virtual environments: a presence
questionnaire. Presence Teleoper. Virtual Environ. 7(3), 225–240 (1998)
Tirolcraft: The Quest of Children to Playing
the Role of Planners at a Heritage
Protected Town
Bruno Amaral de Andrade(&), Ítalo Sousa de Sena,

and Ana Clara Mourão Moura
Geoprocessing Laboratory, School of Architecture,

Federal University of Minas Gerais, Belo Horizonte, Brazil
{deandradebruno,italosena,anaclara}@ufmg.br
Abstract. The main goal of the article is to explore the potential of Minecraft
as a platform to engage children into participatory planning. The game enables
the players to easily design using blocks to build structures like houses, play-
grounds, lakes, vegetation, agriculture, etc. The area of study is a town called
Tirol, a heritage protected settlement built by austrian immigrants in the
municipality of Santa Leopoldina, State of Espírito Santo, Brazil. This article
advances on the state of the art by articulating the potential of Minecraft as a
game-based learning into urban participatory planning with children as pro-
tagonists actors of rethinking the city. Also, the game enables children to design
appropriating themselves on the concept of “child-friendly city” and discussing
their design ideas with each other collaboratively. The results indicate that
children can learn and work on a playful way to collaborate on urban planning
processes, and widens open new researches possibilities.
1 Geogaming: Serious Games & Urban Planning
The development and application of Serious Games in Urban Planning are on its first
steps, but have been through an exponential growth as means to help citizen enhance
competences like playing the role of designers of their city. The central reference on the
hybrid theme “Serious Games & Urban Planning” is Alenka Poplin1, whose researches
and publications helped and guided us develop our own research structure, like the new
concept of Geogames (Vemuri et al. 2014).
Hence, this article aims to explore the potential of Minecraft as a serious gaming
platform to acquire knowledge to engage civic participation on redesigning the place,
especially giving the children a leading role. The participatory design was conducted
through an intuitive psychogeographic input of non-expert users, and as a digital
instrument we used an existing game to get straight to the planning phase.
1
Assistant Professor, Department of Community and Regional Planning, College of Design, Iowa
State University. Research interests: GeoGames for urban planning & Serious games for civic
engagement, User experience with interactive maps & GeoVisualization.

DOI: 10.1007/978-3-319-48496-9_66
826 B.A. de Andrade et al.
With the support of free and open source software World Painter2 and MicroDEM3
using real world data of the study area to analyze and visualize a territory model to
export to Minecraft. Then, in Minecraft there was an opportunity to provide learning
processes and rapid meaningful feedback. This experience was an opportunity to come
up with important analyzes in co-creating personalized and pervasive games for
planning.
Therefore, elects an empirical approach as a concrete-object, the town of Tirol, in
Santa Leopoldina, State of Espírito Santo. First, we built a digital elevation model of
the territory and then make it thematic with anthropic and landscape elements. With the
modeling of a real case study on a virtual platform, seeks an empathic relationship of
the children with their territory so that they feel engaged to perform specific inter-
ventions, where their range of scale reaches, for example, the home to school route, and
other routes and places they visit constantly (Poli 2006).
Furthermore, this article also seeks on a broader perspective come up with relevant
questions that will be pursued during the sequence of the research in geogames, such
as: How could geogames better work on benefiting citizen on knowledge and aware-
ness about the territory? Serious games as a game-based learning can really be a more
attractive instrument to participatory design? What are the stories of games that best fit
children’s participation in urban planning? What kind of a pilot geogame be like to
reach and attract more engagement and participation of social actors?
2 Serious Games and Geogames
Vemuri et al. (2014) expatiating on the differences between the concepts of games,
serious games and geogames. The last one that interests us the most for this article is
conceptualized as serious games focused on learning about the territory and the dif-
ferent ways of representing and visualizing it in a game-based environment. The
authors refers by the same tolken of the concept of geogames to Ola Ahlqvist4 and
Christoph Schlieder5, who conducts researches on geogames, extolling the movements
of players involving locomotion and thereby the physical effort characteristic of any
sportive activity as a missing element in interactive console games. Interestingly,
Schlieder et al. (2005, p. 168) defines what is not a geogame, that is “(…) games that
do not satisfy the spatio-temporal coherence constraints in which resources magically
jump around the board – are not geogames”.
2
Available on http://www.worldpainter.net/, accessed in 20 May 2016.
3
Available on http://www.usna.edu/Users/oceano/pguth/website/microdem/microdem.htm, accessed
in 20 May 2016.
4
AHLQVIST, Ola; SCHLIEDER, Christoph. Geogames and Geoplay Game-based Approaches to the
Analysis of Geo-Information. “GeoGames – a virtual simulation workbench for teaching and
learning through real-world geography”. Available on http://geogame.osu.edu/, accessed in 18 May
2016.
5
Assistant Professor of Computing in Cultural Geosciences, Head of the Geogames Team, Laboratory
for Semantic Information Processing, Otto-Friedrich-University Bamberg, Germany.
Tirolcraft: The Quest of Children to Playing the Role of Planners 827
Basically, the goal of geogames is to be one or a series of decision-making

game-based models capable to support participatory processes in urban planning (Tóth
and Poplin 2014).
3 Why Use Minecraft as a Game-Based Learning Tool?
Until August of 2016 Mojang, company that developed the game, has sold more than
24 million Minecraft (PC version), what shows the importance of this game in the
world scene. This success is primarily related to the possibility of the player interact
with digital world on an experience to build or destroy anything. This characteristic
also make this game an endless one, leading to an infinity of possibilities. The block
logic brings to the player an experience of endogenous connection to learn, when the
environment drives the player to explore the territory to gather supplies from the nature
and build structures like houses, castles, mines, vegetation, etc. These aspects of the
game have an intrinsic potential to a game-based learning, because of the way that the
player interact with the environment.
Colin Gallather organized in 2015, with others teachers, a manual that shows how
Minecraft could be used in classrooms. They explain that this game develop concepts
like creativity, collaboration (in multiplayer online servers), digital citizenship, fun,
leadership, differentiation, engagement, independence, also others aspects of the
learning process. Sean Dikkers (2015) published a book called Teachercraft, where he
explains the possibilities of applying Minecraft in educational approach, exploring the
usefulness of the game in the knowledge-based students’ formation, but also in the
approximation of the teachers with the digital environment. This potential shows that
Minecraft also can be used as a tool to develop others topics by the flexibility of the
game structure logic.
4 Why Tirol in Brazil to Play a Participatory Game

with Children?
The reason of choosing Tirol as the study area is in the fact that the municipality of
Santa Leopoldina be a territory of heritage interest, for having an architectural and
urban perimeter protected by the State Council of Culture, and farms buildings of
sugarcane and coffee cultures from the nineteenth century in the rural areas, such as
Tirol. It is done both a justification through analysis of historic cartography mapping,
referring to the territorial cycles that shaped the palimpsest of Santa Leopoldina, and
the emergence and consolidation of immigrants occupying cores not Lusitanian, in
particular Austrians in Tirol (Andrade 2012).
From the standpoint of the cartography analyses (Fig. 1), the black color highlights
the area of the Imperial Government’s project delimitation called Colony of Santa
Leopoldina, and the stars in yellow demarcate and reveal the names of the countries of
origin of immigrants, from top to bottom, there is Luxemburgo, Pomerania, Tyrol,
California e Hollanda. In fact, the urban center of Santa Leopoldina is outside the area
delimited for the colony, thus raises the hypothesis that the city center is built to give
privilege to the Luso-Brazilian residents for commercial uses, while rural areas were
reserved to non-Lusitanian immigrants as specified above. This hypothesis can be
confirmed in the contemporaneity according to the architectural legacy and diffuse
settlement in the region, related to the former colony (now splitted in various munic-
ipalities, one of them still keeps the name of Santa Leopoldina), so that the immigrants
cultural heritage is located mainly in rural areas, just like Tirol (Andrade 2015).
Fig. 1. Tirol’s location (in yellow) in the municipality of Santa Leopoldina, State of Espírito
Santo, Brazil. (Color figure online) Source: Adapted from Planta da Parte da Província
Espírito-Santo, 1978 (Andrade 2015).
Data updated by the IBGE6 (Instituto Brasileiro de Geografia e Estatística) shows

that the municipality of Santa Leopoldina has an estimated population of 12,885
inhabitants, in 2015, a territorial area of 718.097 km2, and a population density of
17.05 inhabitants/km2. According to the IBGE 2010 census, 21.5 % of the population
lives in the urban center, and 78.5 % live in the rural area (Fig. 2).
This reasoning justifies the interest of this study in the town of Tirol, based on the
assumption of its notorious landscape, its singular history, its citizen, and its
now-a-days problems like the rupture of cultural values and consequently a population
exodus (Andrade 2015). That are, therefore, why we chose to work with the children as
social actors who has the potential to be empowered to Geodesign (Steinitz 2015)
alternative futures to solve the struggles of Tirol.
6
Source http://cidades.ibge.gov.br/painel/painel.php?codmun=320450, Accessed in 11 July of 2016.
Fig. 2. Tirol’s location. Source: Andrade (2015, p. 54)
5 Tirolcraft: Children’s Participatory Design
The activity is developed at the Tirol School for children from pre-school until the fifth
year of the fundamental school, about 04 to 11 years old. We worked with all the
students, which are 30 students divided in two groups: the pre-school and the funda-
mental school. The methodology steps followed: (1) Presentation of the Minecraft
game and its functionalities; (2) The objective of the activity: “Redesigning Tirol”;
(3) One by one every student played the game an put one’s own spin on it; (4) Final
discussion of the projects and what could really be done in Tirol.
The step 1 of the method we designed in the game some structures using blocks, we
showed how the game works and the principal functions of using it on a computer,
which is different from when you play the game on a video-game console. The step 2
consisted in explaining carefully the main objective of this experiment that is
Redesigning Tirol as they would like to see it become. We explained that they could
relate some structures as it is presented now, or was settled historically, or they could
propose something new. The step 3 allowed every student to play the game and build
something ones though it could be built in Tirol to make a better place to live, always
trying to show them how important it is to think and design collectively to reach a
maximum consensus as it is possible. The step 4 we provoked a discussion about the
results of their interventions in the place, what they really thought it could be built and
how the community would have liked it. In addition, we asked which one of the
designs made would be their priority, so that they could start mobilizing the community
to start a movement to bring the game to reality.
The first group was the pre-school one, composed by 12 students. Surprisingly they
have shown a good exploitation of the game, since the attention to the initial presen-
tation and as long as the activity flowed. Which makes us think that we really cannot
judge a younger child to reaching the goal of a geogaming activity, although the
territorialist school tells us to work more with children between 9 and 12 years old (Poli
2006) (Fig. 3).
Fig. 3. Pre-school children playing Minecraft
In the meantime, of the development of the activity, the pre-school children got
distracted, because of the availability only of the researcher’s notebook. The five
computers that the informatics laboratory has only one works but it does not have the
minimum capacity to play Minecraft, and the other four are new but not working.
Besides, we had to play offline, because the school does not have internet anymore, and
the one they had the professors told us it had a bad speed quality (Fig. 4).
Henceforth, some of the projects that the pre-school children projected were
(Fig. 5): (a) buildings with materials coming from the local environment, like wood
and stone; (b) playgrounds with equipment for playing activities; (c) lakes; (d) cavern;
(d) square in front of the school and the church, with lake, trees, flowers, animals, and a
child playground.
Fig. 4. Children getting distracted by the fact of having only one computer to play
Fig. 5. Minecraft design made by pre-school children

The second group was the fundamental-school one, composed by 18 students. They
have shown a desirable exploitation of the game, and have made some very interesting
suggestions, collaboration to each other ideas and most importantly designs. They did
not show any issue on non-concentration like the pre-school ones.
Henceforth, some of the projects that the fundamental school children projected
were (Fig. 6): (a) new buildings like a bigger school; (b) vegetation trees, and gardens
with flowers; (c) cavern; (d) a new square in front of the school and the church, with
lake, trees, flowers, animals and child playground with a places for bicycle, roll-
erblades, kick scooter, and skate (Fig. 7).
Fig. 6. Fundamental-school children playing Minecraft
6 Final Considerations and Further Research Steps
First of all, it is incredible how children can quickly learn a game, by its inherent
motivational aspect of being fun, other than being an innovative method of teaching at
school, the fact that their own territory is modeled urging an affective reaction con-
cerning the place as it matches their own mental maps, and the empowerment aspect of
decision about future scenarios for their town.
Notwithstanding, some of the ideas of the Tirolcraft experiment remained only on
the discourse, like the will to design paved streets and sidewalks, a big square to reunite
the community, a shopping mall, and an amusement park. Speaking of that, in Tirol
Fig. 7. Minecraft design made by fundamental-school children
there are almost no pavements, most of it are still on dirt roads with no regular
maintenance. Which could be explained that as much as they would take the game
seriously on a specific purpose, the fun aspect, the innovation activity at the school, and
the fact that most of them never played Minecraft were distracting for them.
On a broader perspective, the development and application of this methodology,
technology and model aims to favor cultural landscape maintenance, both by identi-
fying heritage values and developing civic consciousness. This first step of the
approach is to strengthen the involvement of the children, to listen and hear them, as a
way to approximate to the whole community. Then it becomes possible to the next step
on a game-based learning approach to rebuild digitally the cultural landscape and lead
the community to identify values for furthering guidelines and best practices related to
spaces for protection and planning. Hence, this second step relates to the following
works in progress provisionally entitled “3D Virtual Cultural Landscape Navigation of
Tirol” and “The Sticker Album of Tirol”.
Concerning the questions that moved us to this experiment, geogames appears to
better work, both from this experiment with children and from the lessons learned
specially from Alenka Poplin, when it has a playful environment with a tight objective
focus. Alongside, game-based learning could be a powerful instrument to be used on
participatory design regarding its engaging flavor. The stories to be told on a game
seems to be on the direction of best representing and visualizing a real world case
study, with elements that come from mental maps of the local community.
Overall, it was a successful experiment since the modeling using mostly free and
open source GIS software, although some adjustments had to be made like fixing from
inside Minecraft the smaller parcel of the center town, the location that they have in
domain in their mental maps. The children showed and proved to be relevant social
actors to be considered in urban planning processes coming up with ideas, even
detailed ones, from a delicate perspective that could enrich a participatory design of a
place.
Finally, a geogame pilot that could be developed to reach and attract more
engagement and participation is one capable of engaging and motivating the citizen on
a real world tridimensional modeling. Adding the most elements as possible from
mental maps of its own community, like even sounds of the place, real people dia-
logues, little quests to best known and recognize the place. The format should be one
played on various digital devices, with the social actors online being able to talk to each
other and find a consensus to reach at the end of the process alternative futures for their
city.
References
de Andrade, B.A.: Uma rota patrimonial para o rio Santa Maria da Vitória, como instrumento de
conservação, valorização, requalificação e/ou transformação do patrimônio territorial.
Graduation Thesis, Department of Architecture and Urbanism, Federal University of Espírito
Santo, Vitória, Brazil (2012)
de Andrade, B.A.: Representando o Patrimônio Territorial com tecnologia da geoinformação.
Experimento em Santa Leopoldina, Espírito Santo, Brasil. Master Thesis, Post-Graduation
Program in Architecture and Urbanism, Federal University of Espírito Santo, Vitória, Brazil
(2015)
Poli, D.: Il bambino educatore: progettare con i bambini per migliorare la qualità urbana, p. 270.
Alinea Editrici, Florence (2006)
Poplin, A.: Games and serious games in urban planning: study cases. In: Murgante, B., Gervasi,
O., Iglesias, A., Taniar, D., Apduhan, B.O. (eds.) ICCSA 2011. LNCS, vol. 6783, pp. 1–14.
Poplin, A.: Digital serious game for urban planning: “B3 – Design your Marketplace!”. Environ.
Planning B Planning Des. 41(3), 493–511 (2014). Pion Ltd., London
Poplin, A.: Playful public participation in urban planning: a case study for online serious games.
Comput. Environ. Urban Syst. 36(3), 195–206 (2012). Elsevier
Schlieder, C., Kiefer, P., Matyas, S.: Geogames: a conceptual framework and tool for the design
of location-based games from classic board games. In: Maybury, M., Stock, O., Wahlster, W.
(eds.) INTETAIN 2005. LNCS (LNAI), vol. 3814, pp. 164–173. Springer, Heidelberg (2005).
doi:10.1007/11590323_17
Steinitz. Um framework para o Geodesign. Alterando a Geografia através do Design, p. 208. Esri
Press, California (2015)
Tóth, E., Poplin, A.: ParticiPécs – a cooperative game fostering learning about the built
environment and urban planning. In: 17th AGILE Conference on Geographic Information
Science (AGILE), Workshop Geogames and Geoplay, Castellón, Spain (2014)
Vemuri, K., Poplin, A., Monachesi, P.: YouPlaceIt! a serious digital game for achieving
consensus in urban planning. In: 17th AGILE Conference on Geographic Information Science
(AGILE), Workshop Geogames and Geoplay, Castellón, Spain (2014)
Pervasive Game Utilizing WiFi
Fingerprinting-based Localization
Filip Maly, Pavel Kriz(B) , and Michael Adamec
Faculty of Informatics and Management,

Department of Informatics and Quantitative Methods,
University of Hradec Kralove, Hradec Kralove, Czech Republic
{Filip.Maly,Pavel.Kriz}@uhk.cz
Abstract. The ability to find out a geographical position of a user is

one of the unique features of today’s mobile devices. The aim of this
work is to suggest and implement a pervasive game for the Android
operating system which will utilize two methods of the localization of
the mobile device simultaneously. The application should guide the user
through historical sites and other places of interest in the town. It could
increase their attendance as well. The combination of the two ways of
the localization will allow us to achieve two goals. First, to verify if the
user is really located at the given place (i.e. that he/she did not mock
his/her position via Developer Options at the Android system). Second,
to create and update our own database of WiFi fingerprints usable for
faster WiFi-based localization.
Keywords: Localization · Fingerprinting · WiFi positioning · Geofenc-

ing · Gamification · Pervasive game · Android · GPS
1 Introduction
A mobile phone is no longer a tool for making phone calls and sending SMS
messages only. Similar to human beings, today’s mobile phones are able to receive
and process information from the surrounding environment. Various types of
sensors have been added to the mobile phones.
A sensor for determination of the geographical position is one of the most
important sensors today. Most devices contain a chip able to receive a signal
of GPS (Global Positioning System). The Assisted GPS (A-GPS) technology is
often supported by the online devices. To speed up the localization of a device,
A-GPS downloads data necessary for the localization quickly from the Internet
instead of downloading them from the GPS satellites via a slow data-link. Despite
this, the time-to-first-fix of an A-GPS receiver may vary from approx. 10 to 30 s
which may be a frustrating period for users [1].
Information about the user’s geographical position is very valuable for devel-
opers of mobile applications. These data can be used for navigation on the map,
for suggestion of the nearest bus stops when looking for transport connection,
for sharing the position with friends at the social networks, etc. [2,3].

DOI: 10.1007/978-3-319-48496-9 67
Pervasive Game Utilizing WiFi Fingerprinting-based Localization 837
The main goal of this work is to suggest and implement a pervasive game
for the Android operating system which will use more options to determine
the geographical position of the mobile device and to test the feasibility of the
solution. The application will show the most famous Places of Interest (POI) in
the town guiding the user to them. The user will be assigned certain amount
of points when he/she visits the POI. The players will then compete with other
participants to achieve a higher score. At the same time, the application will
enable to synchronize the user’s game profile among multiple devices and a web
portal. Thus, the user will be able to log in at any of his/her mobile devices
and will always have the up-to-date data about the places visited and about the
amount of points assigned. It will also be possible to provide special offers (for
example a reduced entrance fee) to the users based on their score or their choice
of the previous POIs (personal preferences). It can increase attendance of more
POIs involved in the game.
Our work will deal with the issue of the multiple-source localization of devices
in the Android operating system, geofencing and synchronization of the profile.
The rest of this paper is organized as follows. We formulate the problem in
Sect. 2. Section 3 describes the existing positioning techniques and related work.
In Sect. 4, we analyze the requirements of the proposed application. Section 5
describes the proposed solution. Several details regarding the implementation
are shown in Sect. 6. We present the results of the testing in Sect. 7. Section 8
concludes the paper.
2 Problem Formulation
The main objective of our mobile application is to guide the user on the map
to the historical site or another POI. Then, the application will verify if he/she
really is at the given place. This verification will be done in two ways – using
GPS and using fingerprints of WiFi networks. Owing to this, it will be possible
to find out if the user is or is not cheating, for example with the aid of location-
mocking enabled by the Developer Options in the Android operating system.
Thanks to this solution, our own database of WiFi networks fingerprints will
also be filled. This database will serve for the WiFi-based localization that is
faster than A-GPS and improves the User Experience (UX) thus.
The application will consist of two parts. The first part is a web-based appli-
cation interface (API) which will enable to download information about POIs
into the device, to log into the service, to synchronize the user’s profile and to
compete with the other users. The second part is the mobile application for the
Android system. The Android platform has been chosen because it is the most
wide-spread mobile system at the market [4].
3 Positioning Techniques and Related Work

3.1 Positioning Using Cellular Networks
The positioning based on cellular networks is the least precise, but on the other
hand the most energy-efficient method. To localize the device, the information
838 F. Maly et al.
about the communication of the mobile phone with the Base Transceiver Stations
(BTS) are used. Currently, the trilateration (multilateration in general) is often
used [5]. Depending on the particular solution and the density of base stations,
its accuracy is about hundreds of meters. It is necessary to know the exact
geographical position of all base stations of the cellular network. The localization
based on cellular networks is available through the LocationAPI service. There
are also free alternatives, such as the OpenCellId service [6].
3.2 WiFi-Based Positioning
The positioning based on the surrounding WiFi transmitters is another option

for the localization of the mobile device. This way of the localization is the most
accurate in cities with a large number of WiFi networks.
Each WiFi access-point (AP) has its own MAC address. The principle of
the localization itself is very similar to the determination of the position using
the cellular networks. The mobile phone is able to find out the received signal
strength (RSS) and the MAC addresses of the individual APs of the surrounding
WiFi networks.
The calculation of the position itself can be done in different ways depending
on the method used [7]. Multilateration and fingerprinting are the main methods.
As for multilateration, it is necessary to know the position of the particular APs.
As for fingerprinting, it is necessary to build a database of fingerprints in advance
and then to search for the most similar (or k of the most similar ones) fingerprint
with the known position. In the ideal case, the localization using WiFi can have
accuracy of meters at the cost of very low energy consumption [8].
3.3 Bluetooth-Based Positioning
Bluetooth transmitters can be used for the localization in a similar way as the
WiFi access points [9,10]. Due to the limitation of the original Bluetooth spec-
ification (such as a lengthy process of the device discovery), this approach has
not been used much. The situation changed with the advent of Bluetooth 4.0 in
2010. Due to low energy consumption and fast device discovery, the utiliza-
tion of this technology is much more promising [11,12]. But still, there are not
as many Bluetooth transmitters (called beacons) in the public space as WiFi
access points.
3.4 Global Navigation Satellite Systems

GPS is the oldest public Global Navigation Satellite System (GNSS), with the
aid of which it is possible to determine the position and the precise time anywhere
on the Earth. The GPS can be divided into three segments.
The space segment currently consists of 32 orbiting satellites [13]. The Control
segment monitors and controls the cosmic segment of the system and updates
the satellites’ movement model [14]. The user segment is the last part. Users
receive signals from the individual satellites with the aid of the GPS receiver.
The receiver can derive its position based on the time of flight (TOF) of the
signals (corresponding with the distance among the satellites and the receiver).
The accuracy of the publicly available part of the GPS may vary from meters
to tens of meters, depending on many factors [15].
3.5 Related Work
The WiFi-based localization is not a novel idea. RADAR [16] is one of the first
systems based on WiFi fingerprints. This topic becomes hot in recent years due
to widespread smart-phones [8] and ubiquitous location-based services. Most
authors focus on indoor localization [17], because GPS usually fails indoors.
There are also location-based mobile applications with gamification aspects
such as Foursquare. Unfortunately, they are not proof against cheating because
they use a common location service on the mobile device, which can provide
mocked coordinates.
In contrast to the existing solutions, we present a novel approach to check
the presence outdoors using the cross-verification between the GPS-based and
the WiFi-based localization.
4 Analysis
In the previous section, we have described the ways of the localization of a mobile
device outdoor. Each way has its own advantages and disadvantages. While the
cellular-network-based localization is the most energy efficient, is not suitable for
navigation in the pervasive game because of the insufficient accuracy. The local-
ization based on WiFi has an improved accuracy while requiring the database
of access-points or their fingerprints near points of interest at the beginning of
operation. It would also be required to update the database because of changes
in the wireless networks. The localization using GPS is the last option. This
method is precise enough for navigation on the map and for decision if the user
had really visited the given place. Its drawbacks are the longer time necessary
for carrying out the localization, higher consumption of the electric energy and
the possibility of relatively easy faking of the GPS coordinates by the user. It
could happen that the user could get points for visiting POIs without physically
being there.
5 Proposed Solution
The aim of the application is to guide the user at for example historical sites and
other interesting places in the city. Navigation built into the Google Maps will be
used in the guidance phase. We have designed the following work-flow. First, the
player logs into the application. Then, the list of the cities is offered to him/her,
from which he/she can download one to the mobile phone using the web API.
840 F. Maly et al.
Thanks to this initial phase, a large amount of data can be downloaded via WiFi
avoiding speed and data limits forced by the user’s cellular data plan. Later on,
the mobile connection is used only when “activating” the place (the sight) itself
by the user, for downloading the current leader-board and for displaying the
map (if not cached). Nevertheless, the volume of the data transferred can be
minimized by the use of cache.
5.1 Verification of the Presence
We have to be able to verify the user’s physical presence at the place prior to
the assignment of the points for visiting it. Of course, it is possible to check the
distance from the place based on the GPS coordinates only. Then, if the user is
close enough, he/she can acquire the corresponding number of points. But there
are Android applications that can mock the GPS coordinates reported to other
applications, especially for rooted devices. In this way, the user could get points
(and potential rewards) for visiting a place without actually being there. This
issue is solved by scanning the surrounding WiFi networks and creating a so
called fingerprint at the given place.
It is necessary to scan the surrounding WiFi networks (including their MAC
addresses and RSS) near the POI before filling the data about the POI into the
database. This initial phase can also be performed by scanning WiFi fingerprints
additionally by the application’s users themselves. This is a certain form of
crowd-sourcing. The scanned fingerprints from the first users are used as a base
for further verification. Later fingerprints may keep the database up-to-date
when cross-verified among multiple users at the same place.
The procedure of verification of the presence at a given POI is the following:
after activating the place by the user, the distance is checked using the GPS
coordinates. This distance has to be shorter than the distance in the database
which is defined individually for each place. Than, the presence of the sufficient
number of WiFi networks with the significant RSS at the particular place is ver-
ified. If there were not enough significant WiFi networks near the place during
the scanning, the points for visiting the place should be assigned directly. This
situation has to be always verified by the data from more users. If the user does
not receive any networks at a given place and the others do, it is a highly sus-
picious situation. If a sufficient number of networks is received, the surrounding
wireless networks are compared with those in the database using their MAC
addresses. In this way it is verified if the user is really present near the given
place. Then, the corresponding number of points is assigned to him/her.
5.2 Profile Synchronization
The user should be allowed to play the game on any of his/her devices and
should always see the up-to-date data at it. This especially means to have marks
at the already visited places and the number of points that he/she has got for the
visiting. Thus the synchronization of the user’s profile has to be implemented.
To send the profile changes from the server to the mobile device using Push
Notifications, we could utilize the Google Cloud Messaging (GCM) [18]. But
finally, our own way of synchronization technique has been designed. During the
logging into the application, the unique API key is generated for the user. If
the user logs in the application from another device, another API key will be
generated. The API key acquired for the first device becomes invalid at that time
and the application logs out automatically. This means it is possible to play the
game on one device only in a given moment. In this way it is ensured that it is
enough to download the user’s profile data at login-time only and they do not
have to be synchronized continuously while the game is played on the particular
device.
6 Implementation
The application consists of two parts; server-side and client-side. The server-
side is written in the PHP scripting language and the data are saved in the
MySQL database. Figure 1 shows the database model. The REST architecture
(Representational State Transfer) has been chosen for the API [19].
Fig. 1. Database model
The client-side application is supposed to run at the Android version 4.0.3

or later. The application utilizes several libraries; the Support Library that
ensures the compatibility with older versions of the Android operating system,
the Picasso library handling the pictures, the Retrofit library with the OkHttp
HTTP client for communication with the web API, and the Material-ish Progress
library.
6.1 Localization Implementation Details

For the GPS-based localization, we use the Fused Location Provider at the
Android platform. We also use the geofencing feature via the GeofencingAPI. It
842 F. Maly et al.
may fire an event when the user enters or leaves a certain defined area on the
map. Each POI is assigned with the latitude, the longitude and the radius of the
monitored area. If the user enters this area, the button appears on the screen
with the aid of which it is possible to activate the POI by the user and then
to get points for visiting it. The GeofencingAPI is not able to add more than
100 monitored areas [20]. But this was taken into account in the design of the
application and individual POIs are distributed grouped by the towns in order
to circumvent this restriction.
6.2 Graphical User Interface
The user interface is localized into Czech and English language. Major compo-
nents of the Graphical User Interface (GUI) will be described in the text.
In the Fig. 2a, the main screen of the whole application with the game map is
shown. On the map there are red spots representing the unvisited POIs. There
are transparent blue circles around the unvisited POIs indicating the area where
user can activate the POIs and receive the points. The detail of the POI (see
Fig. 2b) contains the information whether the POI has or has not been visited
(activated), the distance from the POI, the number of points which will be
assigned to the user for activation and the text with the basic information about
the chosen POI. In the Fig. 2c, there is a list of POIs. If the valid position of
the device is available, then this list can be sorted according to the distance to
individual POIs.
(a) Main screen (b) POI details (c) List of POIs nearby
Fig. 2. Screenshots
7 Initial Measurements
Testing of the application was divided into several phases. In the first phase, the
presence of a sufficient number of wireless networks with an acceptable strength
of the broadcasted signal near individual sights (POIs) had to be found out.
Seven most important sights in Hradec Kralove, Czech Republic, were chosen
based on [21]. Near these sights, thorough measurement of surrounding wireless
networks was done. The results in the Table 1 show that there are more than 10
wireless networks near majority of sights in the town. The St. Nicholas Church
is the only place where the number of networks is extremely low having low RSS
values. Thus, this place is not suitable for the proposed cross-verification.
Table 1. Number of wireless networks and their signal strengths around the chosen
sights in Hradec Kralove, Czech Republic
Name of the sight # of WiFis Mean RSS [dBm] Median RSS [dBm]
White Tower 10 −88,50 −89,0
Holy Spirit Cathedral 14 −88,43 −90,5
Church of the Assumption of the Virgin 19 −91,42 −92,0
St. Nicholas Church 3 −92,67 −94,0
Hucak power station on the Elbe River 12 −91,01 −92,0
Plague Column of the Virgin Mary 19 −89,16 −91,0
Muzeum of Eastern Bohemia 37 −90,65 −92,0
Figure 3 shows a histogram of the particular AP’s RSS values in the White
Tower location. During comparison of fingerprints, one should consider using a
filter that ignores unstable APs with RSS lower than a threshold value. Note
that different devices may report different RSS values. This should be taken into
account introducing a normalization function that will be device-specific. Based
on our measurement, RSSs lower than 92 dBm (measured with Samsung Galaxy
S4 Mini) are reflected and/or unstable.
Fig. 3. A histogram of an AP’s RSS values in the White Tower location

844 F. Maly et al.
8 End-to-End Tests
In the second phase of testing, the application was installed to five users’ mobile
phones for a detailed qualitative testing based on one-on-one interviews. All
users were asked to log in the application, download the POIs in Hradec Kralove
and visit at least one sight. During this trial, several minor issues were revealed.
After log-in, the list of towns available to download is displayed. Based on
the users’ responses, the dialog warning against the excessive data download has
been added. Then, the timeout regarding the communication with the server
had to be adjusted from 10 to 20 s. After elimination of these issues, the users
were satisfied with the application. They appreciated its graphic arrangement,
low consumption of the mobile data and overall speed of the response.
9 Conclusion
In this paper, we have described the design and implementation of the novel
Android application for tourists having unique features; the fingerprint-based
localization and the location verification. The application is currently in closed
beta testing but the APK is available1 . The aim of the application is to guide
the user to chosen places of interest (e.g. historical sites) in the town and to give
him/her the interesting information about them. The application could also con-
tribute to a higher attendance of these places by incorporating the gamification
approach.
The use of the WiFi fingerprinting for validation of users’ presence at the area
of the POI is the main contribution. In contrast to other similar applications,
the fingerprints of wireless networks are used for localization besides the GPS.
Thus, the user is prevented from manipulating his/her geographic coordinates
on the map and from cheating. The fingerprint database may also be used for the
faster localization at the POI before the GPS receiver reaches the time-to-fix.
The presented approach represents an effective way for WiFi data collecting.
The application was thoroughly tested in different versions of the Android.
For the real deployment it is necessary to create a sufficiently extensive database
of places so that the application is as attractive for the users as possible. In future
work, we will focus on incorporating other technologies such as the promising
Bluetooth Low Energy Beacons into the solution.
Acknowledgements. The authors of this paper would like to thank Tereza Krizova
for proofreading. This work was supported by the SPEV project, financed from the
Faculty of Informatics and Management, University of Hradec Kralove.
1
http://edu.uhk.cz/∼krizpa1/doku.php?id=geocity.
References
1. Zhang, J., Li, B., Dempster, A.G., Rizos, C.: Evaluation of high sensitivity GPS
receivers. In: International Symposium on GPS/GNSS, Taipei, Taiwan (2010)
2. Singhal, M., Shukla, A.: Implementation of location based services in android using
GPS and web services. IJCSI Int. J. Comput. Sci. Issues 9(1), 237–242 (2012)
3. Vanjire, S., Kanchan, U., Shitole, G., Patil, P.: Location based services on smart
phone through the android application. Int. J. Adv. Res. Comput. Commun. Eng.
3(1), 4982–4987 (2014)
4. IDC: Smartphone OS market share, 2015 Q2 (2016). http://www.idc.com/
prodserv/smartphone-os-market-share.jsp
5. Kan, K.K.H., Chan, S.K.C., Ng, J.K.Y.: A dual-channel location estimation system
for providing location services based on the GPS and GSM networks. In: 17th
International Conference on Advanced Information Networking and Applications,
AINA 2003, pp. 7–12, March 2003
6. Trevisani, E., Vitaletti, A.: Cell-ID location technique, limits and benefits: an
experimental study. In: Sixth IEEE Workshop on Mobile Computing Systems and
Applications, WMCSA 2004, pp. 51–60, December 2004
7. Honkavirta, V., Perala, T., Ali-Loytty, S., Piche, R.: A comparative survey of
WLAN location fingerprinting methods. In: 6th Workshop on Positioning, Naviga-
tion and Communication, WPNC 2009, 243–251, March 2009
8. Machaj, J., Brida, P.: Using of GSM and Wi-Fi signals for indoor positioning based
on fingerprinting algorithms. Adv. Electr. Electron. Eng. 13(3), 242–248 (2015)
9. Bruno, R., Delmastro, F.: Design and analysis of a Bluetooth-based indoor local-
ization system. In: Conti, M., Giordano, S., Gregori, E., Olariu, S. (eds.) PWC
978-3-540-39867-7 66
10. Muñoz-Organero, M., Muñoz-Merino, P.J., Delgado Kloos, C.: Using Bluetooth to
implement a pervasive indoor positioning system with minimal requirements at the
application level. Mob. Inf. Syst. 8(1), 73–82 (2012)
11. Budina, J., Klapka, O., Kozel, T., Zmitko, M.: Method of iBeacon optimal distri-
bution for indoor localization. In: Christiansen, H., Stojanovic, I., Papadopoulos,
G.A. (eds.) CONTEXT 2015. LNCS (LNAI), vol. 9405, pp. 105–117. Springer,
Heidelberg (2015). doi:10.1007/978-3-319-25591-0 8
12. Kriz, P., Maly, F., Kozel, T.: Improving indoor localization using Bluetooth low
energy beacons. In: Mobile Information Systems (2016). http://www.hindawi.com/
journals/misy/2016/2083094/
13. El-Rabbany, A.: Introduction to GPS: The Global Positioning System. Artech
House Mobile Communications Series. Artech House, London (2002)
14. Seppänen, M., Ala-Luhtala, J., Piché, R., Martikainen, S., Ali-Löytty, S.:
Autonomous prediction of GPS and GLONASS satellite orbits. Navigation 59(2),
119–134 (2012)
15. Modsching, M., Kramer, R., ten Hagen, K.: Field trial on GPS accuracy in a
medium size city: the influence of built-up. In: 3rd Workshop on Positioning, Nav-
igation and Communication, pp. 209–218 (2006)
16. Bahl, P., Padmanabhan, V.N.: RADAR: an in-building RF-based user location and
tracking system. In: INFOCOM, pp. 775–784 (2000)
17. Rai, A., Chintalapudi, K.K., Padmanabhan, V.N., Sen, R.: Zee: zero-effort crowd-
sourcing for indoor localization. In: Proceedings of the 18th Annual International
Conference on Mobile Computing and Networking, Mobicom 2012, pp. 293–304.
ACM, New York (2012)
846 F. Maly et al.
18. Sneps-Sneppe, M., Namiot, D.: Spotique: a new approach to local messaging.
In: Tsaoussidis, V., Kassler, A.J., Koucheryavy, Y., Mellouk, A. (eds.) WWIC
978-3-642-38401-1 15
19. Masse, M.: REST API Design Rulebook. O’Reilly Media, Inc., Sebastopol (2011)
20. Google Android Developers Portal: Creating and monitoring geofences (2016).
http://developer.android.com/training/location/geofencing.html
21. Hradec Kralove Offical Webpages: History and heritage - Significant
Hradec Kralove Monuments (2016). http://www.hradeckralove.eu/tourists/
history-and-heritage
Project Paper: Digital Cultural Heritage
in Education, Learning and Training
ErfgoedApp
An Educational Experiment with Augmented Reality, Cityscapes
and Campusscapes in Brussels
Marc Jacobs1,2 ✉ , Morien Schroyen2, and Joke Vanderschoot2

( )
1
FARO, Flemish Interface for Cultural Heritage, Brussels, Belgium
[email protected]
2
HARP, Vrije Universiteit Brussel, Brussels, Belgium
{mmjacobs,morien.schroyen,joke.vanderschoot}@vub.ac.be
Abstract. In 2015 in Flanders (Belgium) an ErfgoedApp (heritage app) was

launched. It was developed by Vidinoti and FARO with PixLive. The program
allows to construct and use Augmented Reality applications, linked to heritage
items, collections and institutions. In 2015–2016 master students in archaeology,
arts sciences and archivists experimented with the app, as part of the course work.
They managed to produce applications that work, provided feedback to further
develop the App and offered reflection on the relation between heritage work,
cityscapes and augmented reality and the differences between working with or
towards texts or visual information. Low or no cost for distributing and using the
applications in practice in heritage and academic contexts proved possible.
Keywords: Erfgoedapp · Heritage app · Augmented reality · Cityscapes ·

Campusscapes · Heritage studies
1 ErfgoedApp
First we briefly present the ErfgoedApp, an app that is introduced for free in Flanders
not only for users but also for heritage institutions and students, as a tool to generate and
experience Augmented Reality linked to heritage items and representations in the form
of images. Then we will discuss the context in which the tool was launched, the organ‐
ization that was the driving force and the underlying goals. In the third section we will
discuss the context of the courses and some feedback that was provided. In the fourth
part we will present one example and some issues related to it.
What is the ErfgoedApp (heritage app) and how does it work [1]? On the one hand,
for the user (the visitor, the public, the student, …), it is a tool that one can use with an
iPhone or another smartphone or an iPad or another tablet.1 It is first necessary to down‐
load the app (10 MB) from the mayor app stores, via i-tunes store or the play.google.com/
store/apps. This is for free. This has to be done only once: please try it out yourself. The
1
The app requires iOS 7.0 or more recent for iPad or iPhone and 4.0 for phones and tablets that
function on Android.

DOI: 10.1007/978-3-319-48496-9_68
850 M. Jacobs et al.
second necessity is that Bluetooth is activated, a standard feature on smartphones or

tablets. Thirdly, one needs a wifi connection or access to mobile internet. There are
however possibilities to also use the app in spaces where no wifi or mobile internet is
available, e.g. via the option of downloading all the information on the smartphone or
tablet.
What can users do? First it is necessary to make a connection between the app and
the heritage item (or representation of it). This can be done by scanning a QR-code. An
alternative method is by receiving and accepting a signal of an iBeacon. An iBeacon is
a small Bluetooth transmitter powered by a battery (life span: two years, price 40 euros
for the heritage institution). It sends a UUID: a universal unique identifier, triggering a
specific scenario connected to one or more heritage items. It can be attached, via a simple
paste or sticky substance, to the wall, floor or ceiling, to a fire extinguisher, near a
painting, sculpture or window. Or, in an out-door version of the iBeacon, against a wall
or a sign, somewhere in a garden or in front of a monument, etc. The range can be
specified, ranging from a few centimeters to 200 m. It allows for negotiating or forcing
a specific viewpoint, listening distance or perspective of the user. The version that will
be released in December 2016 will also work on GPS-locations outdoors, a powerful
combination with the iBeacon indoors.
Once the signal is received, then the app should be self-explanatory, offering visual
and text information linked to an item, to a collection or a location or land- or cityscape.
Scanning a QR code, or specific images or accepting the signal of an iBeacon allows to
synchronize the content. This can yield the desired effect of augmented reality. This
could be in an interaction with a painting, a photo or another work of art through smart‐
phone or a tablet. But it could also involve the scanning of a picture, a drawing or a page
in a book or magazine, on a postcard or sticker, or on a computer screen, with access to
additional content in the form of a video or audio file or as a text.
For the heritage actor or mediator (a museum, an archive, volunteer organizations,
professionals, students,…) the backend of the system is presented as a free tool to
construct an experience and application for the users. The ErfgoedApp was developed
by FARO. Flemish Interface for cultural heritage, sponsored by the Department of
Culture, Youth, Sports and Media of the Flemish government (in Belgium), in collab‐
oration with a spin-off of the Ecole Polytechnique de Lausanne (EPFL), called Vidinoti.
The Swiss partner developed PixLive Maker, a CMS (Content Management System)
that allows to create and manage augmented reality content. A connection is made with
an image (pdf or jpg, or a QRcode) or iBeacons, and then scenarios can be activated.
PixLive Maker provides a GUI-based scripting interface for building scenarios. The
program allows to monitor the way the app is used (by tracking clicks and iBeacon
activity), to follow-up and adapt the content. The state of the batteries can be monitored
at a distance. The metrics allow to track the traffic and the pushed and demanded content.
These iBeacons can work as a complement to GPS and Google maps, reaching users in
indoor areas, tunnels, basements and spaces beyond the reach of satellite signals. Since
2016 the use of multiple languages to present and consult the content is available, a must
in the Belgian context.
ErfgoedApp 851
2 Interface, Brokers and Mediators
The Erfgoedapp is developed and distributed by FARO. In particular FARO’s talented

ICT specialist and project manager Bram Wiercx, is also providing basic and advanced
training in museums, archives and other heritage institutions in Flanders. Learn to DIY
(do it yourself) is the moto of those courses but also the title of a (regularly updated)
self-study module in Dutch that FARO has been developed and made available for free
online [2]. Via erfgoedapp.be experiences are shared, new launches or use in exhibitions
are announced. Important is the feature of www.erfgoedkaart.be. This was a result of a
massive survey of the whole cultural heritage sector in the so-called Prisma project in
2011–2012. More than 1700 museums, libraries, archives, heritage cells or associations
and volunteer organizations in Flanders are presented on Google Maps, with data and
coordinates of those actors. They have priority access to the courses, are attributed a
login on simple request and can obtain iBeacons to use.
The Cultural Heritage Decree (2008, updated 2012, to be revised in 2016) of the
Flemish Community explicitly gave the task to FARO as a Flemish Interface for Cultural
Heritage to inspire, propose, develop and support measures, networks and programs for
museums, archives, libraries, centres of expertise, volunteer organizations, and so on.
FARO also gives advice about safeguarding policy and practice for local, regional and
(inter)national civil servants, politicians and other policy makers. FARO is called a
“steunpunt” (pivot), an institution is up to now located and mediating between the
government, the academic world, international networks and the field of cultural heritage
(museums, archives, associations, networks,…): an in-between and thus interestingly
ambiguous position. “Cultural brokerage”, “mediation” or “boundary spanning” are
concepts used by heritage workers in FARO to describe their individual functions and
“interface” the word for the combined effect. Underlying are theoretical and methodo‐
logical choices, that are related among others to actor-network theory, participatory
methods, reflexivity and the emerging paradigm of critical heritage studies. Cultural
brokerage is a crucial factor here [3, 4]. Every year FARO organizes the Cultural
Heritage Day in Flanders, on the second Sunday after Easter, which reaches on average
250.000 persons, see www.erfgoeddag.be., also a yearly chance to promote the heritage
app. A tool like the ErfgoedApp illustrates the mission to introduce and translate inno‐
vation, to generate DIY applications at low cost and to invite as many actors and sectors
as possible to collaborate in heritage work and reach a large public or audience, or even
better to invite others to participate or co-create.
3 Context of the Courses with an ErfgoedApp Experiment
In 2015–2016, an experiment was conducted at the Vrije Universiteit Brussel in two

courses: “Heritage and ethnology” on the one hand and “Forum: Contemporary issues
in Art History and Archaeology” on the other, both offered by professor Marc Jacobs,
on the master and the master-after-master level, at which in total 30 students participated.
One third were students of the master-after-master archives sciences and (documentary)
heritage management, usually with a previous training as historian (at the master level).
The rest were students aspiring to become master in arts sciences and archaeology. Six
of these students are older, working students that bring much experience, only one
trained and active as a visual content producer. The participants share the characteristics
that they are of course computer literate but not specialists, highly trained in heuristics
and interpreting historical sources, specialists in retrieving information and critically
assessing and using those texts, images and objects. None of them are programmers. In
both courses the students were challenged to use the ErfgoedApp, to develop an appli‐
cation that will be tested and if possible made available publicly by non-specialists. The
participants agreed that at least first years students arriving on campus in September
would be able and seduced to try it out.
Two types of projects were selected. In the course “heritage and ethnology” the
university campus was used as the site where augmented reality constructions had to be
developed. On the one hand special attention was devoted to works of art on campus,
not only statues and art-work in open-air but also the enormous painting of Renaat Braem
and his wife, spread over five flours and the basement in the central building M, where
the rector and most of the central university services are based. That painting evokes the
evolution from the Big Bang to the 20th century (inspiring one student to find inspiration
in the trailer of the Big Bang Theory series). On the other hand the focus was on that M
building itself, a piece of architecture developed by architect Renaat Braem and classi‐
fied as a protected monument in Brussels. The students developed apps for the (free‐
masons’ and other) symbols at the entrance of the building, the statues of the founders
and the building plans. They also developed apps for the Q building and the Kultuur‐
kafee. The last decades the big aula Q was not only used for academic lectures and
honorary doctorates, but also for music performances of pop musicians and concerts of
classical music, and an annual singing feast of the students. The Kultuurkafee was
famous for performances, wild parties and, in dark periods in the past, also for drug
abuse. It was demolished two years ago and a building trajectory to replace it with new
students homes, cultural spaces and a new Culture Café is going on now. The students
conducted interviews with people who used to work, drink, party or meet there and
integrated that material, together with photos and film footage in the app.
A specific requirement was to use archives and/or oral history. Therefore a close
collaboration with the university archives (CAVA) was organized. They welcomed the
students and helped to find the archival sources. The mixed group of arts students and
archivists saw it as a particular challenge to use scans or original archive material in the
app (even adding augmented reality to the archive material by scanning them and
synchronizing with the QRcodes). This provided new challenges about permissions,
copyright, on how to refer to original sources in a visual application. The choice to focus
on festivities and cultural manifestations led the students to track down archive collec‐
tions on campus that were not yet in the university archives, hence contributing to a win-
win-situation with CAVA. Each student made a functioning app, and a research report
on the story told, the sources found and used and the experience with the app. There
were three groups that decided to make a collective presentation, of the buildings M, of
the in between spaces outdoors and the building Q.
In another course called Forum, in 2016 the theme was “cityscapes”. A number of
lectures on that notion in architecture, urban planning but also cultural studies, on
ErfgoedApp 853
applications on for instance the interpretation of photographs or the representation of

cities and towns in the paintings of Peter Bruegel or the Flemish Primitives were given
as an introduction. Also a two hour introduction was given by Bram Wiercx on using
the ErfgoedApp. The students had to select a location in the city of Brussels and develop
an Augmented Reality experience. An extra challenge for the students that participated
in the Forum course was that they had to integrate and discuss the theory and practical
applications of “cityscape” in a paper accompanying their app. Several students opted
to work on the university campus (in Brussels), as mentioned above. One group worked
on the neighborhood of the Rijksadministratief Centrum in the heart of Brussels, where
also the Nationaal Stripmuseum (national museum for comic books, ranging from Kuie
to Guust Flater) is located. They developed a very nice application to discover several
streets, statues and gardens in the city using tools from the comic books in the app,
including for instance the use of text balloons. Other students decided to work on the
North South train connection (Gare du Midi - North Station). In order to build this
trainline, many houses in the city center had to be demolished: an excellent topic to deal
with in Augmented reality. All these apps will be developed further in the course of
2016 and made available for a larger audience.
The general feeling and feedback of the students was that it took an investment of
time, sweat and frustrations, to get past the first hurdles of using a new program and
experimenting with it. But after many hours of experimenting, the participants were
satisfied that they could realize something that looked professional and that actually
worked in tablets and phones. They could apply the skills they had learned, including
the disposition not to be satisfied with easily available information online, in Wikipedia
or brochures or publications. They tried to discover, use, interpret and present new
material and insights. Some of the participants tried to add several layers and contra‐
dicting versions or interpretations or they attempted to generate eye-opener experiences.
They struggled with the format, e.g. of not resorting to the classical annotation systems
like using footnotes and they found ways around this, by smuggling in solutions via
PDFs and reference systems. They were happy to learn that the classes.
The students experimented with but often finally did not opt for iBeacons. This is of
course also partly due to the fact that they did not operate from within a heritage institu‐
tion or organization, with own buildings and other infrastructure. Most of the participants
opted for augmented reality experiments by uploading jpeg-, png-images or pdfs. They
could easily generate these records or QR codes. What the students particularly appreci‐
ated was the fact that the software suggested appropriate formats or sections when using
images. They also found the next step, connecting “content name” to the images and then
building scenarios a feasible step. The PixLive editor allows to connect scenarios and to
add different kind of buttons. Remarkable is that the art history students repeatedly
decided to use a number of the visual gimmicks available, like the “scratch and win”
option. Most of them used the scratch tool for making the user discover other versions,
other images and variations. It will be interesting to see how the game dimension of the
Scratch, Win or Lose possibility will be used, making random options possible.
The most successful results made a kind of storyboard first, on paper or in a drawing
program before starting to build the app. The promoters of the program claimed that the
application was intended more to make connections with one or a series of objects or
image and connect tours or a story to each of them. The students preferred to build
complex tours and deep layers and scenarios, testing the limits of the program. A popular
choice was the easy to use image carousels that user can swipe away. Here again the
suggestion to carefully think about the order of the images first was emphasized by the
students (chronological or by contrasts, organized in regions, per family or age group, per
type of object, materials or significance, …). The present generation of students had no
problems to work with YouTube-, Vimeo- or DailyMotion-videos (available online) and
MP4-videos made, for instance with a smartphone, iPad or video cameras. Here again the
choice was made to mobilize publicly available free software like http://handbrake.fr.
4 The Project of Morien and Joke
Let us present one example, a project developed in one month time, or better in about
60 working hours each, by two VUB-students in art sciences Morien Schroyen and Joke
Vanderschoot [5]. As participants in the Forum course, they found inspiration in the
lecture and the book by Isabelle Doucet (University of Manchester): “The Practice Turn
in Architecture. Brussels After 1968”. She argued that late-20th (and 21st) century Brus‐
sels was composed of a number of layers and styles and she demonstrated and explained
this on the basis of one very rich photograph of a contemporary cityscape. The students
decided to repeat that experiment, to make a long walk in Brussels and ended up in what
they considered as a suitable location. They opted to develop the idea of cityscape, by
finding a vantage point in the center of Brussels, from a location with a distance of less
than 400 meters of touristic hotspots like Manneken Pis, the Central Market and the
Central Station of Brussels. The location they chose was not one of the places where
foreign or domestic tourists normally come. They selected a view from alongside the
Brussels North-South train axis, near the elite high school of the Sint-Jans Bergman‐
scollege and the graffiti rich skateboard ramp and transjunction near Kapellekerk. They
shot the following panorama shot and went looking for different historical links and
layers (Fig. 1).
From left to right they chose entry points, a new skatepark, the Anneessens tower
and the remnants of the old city walls, the Hoogstraat, the Kappelekerk and the Bruegel
sites. To demonstrate the potential they connected virtual tours to each of the sites. For
the Anneessens tower abundant historical information could be found, both on the
construction itself as about the historical figure of Anneessens. The remnants of the old
city walls linked up with other remaining parts in several parts of the city offered another
challenge. The vantage point of the train north south-connection in the 1950s and 1990s
made it possible to mobilise that story of heavy disturbance of the cityscape [6]. It also
allows to foresee a bridge to other apps of other students focusing on that issue, apps
that were not yet available when writing the paper. The skatepark was constructed as a
result of a commission by Leefmilieu Brussel granted to Recyclart in 2003 and realized
via a participatory project. This opened the space during day time and in the evening for
young people between 14 and 24 years old; a skater generation. All these examples are
documented in other types of source material, hence providing challenges to find consis‐
tency. The fourth part explores the Kapellekerk, that originally was located outside the
ErfgoedApp 855
Fig. 1. Basic scheme for the storyboard of the app developed by Morien Schroyen and Joke
Vanderschoot (Forum course VUB, 2016)
city core but developed as the central point of a vibrant quarter. It was the neighborhood
where the famous artist Pieter Bruegel was located. The internal history and the restau‐
ration of the church proved to be quite interesting.
Joke and Morien made an assessment of the risks of using iBeacons. The chance for
theft or vandalism in the skate park was deemed too high. They negotiated with the
person responsible for the Kapellekerk and obtained the permission to attach iBeacons
in the church but also to consider putting sign or board to offer QR codes. Also leaflets
could be made. They decided to develop a brochure that can be multiplied, and hence
the QR-codes and images.
In the project of Morien and Joke the so-called Bruegelhuis was part of the appli‐
cation. Although there seems to be no hard evidence, the claim is made by Brussels
officials that it was the house where Bruegel allegedly lived during the last days of his
life. In 2019, the 450th anniversary of the death of the famous painter will be the occasion
for big festivities. The Flemish community has opted to turn this into one of its main
events, for attracting tourists, resulting in major exhibitions, connected to the block‐
buster exhibition project in Vienna in 2019. The city of Brussels, but also the Royal
museum of fine arts in Brussels, will also organize activities, focusing on the Bruegel‐
huis. Today, that 16th century house still has to be developed, constructed if one might
say. It cannot be visited yet. In May 2016 Interpret Europe and Herita organized a
congress in Belgium. From the congress venue in Mechelen they organized a number
of workshops in Flanders, including Brussels. One of the workshops was about
“Interpreting Bruegel” in Brussels. On the one hand the Bruegel Box, an installation
powered by Google Cultural Institute in the Royal Museum of Fine Arts in Brussel was
visited and tested. On the other hand the so-called Bruegel house in de Hoogstraat was
visited. Afterwards a workshop took place in the offices of FARO, located nearby.
During the discussions of the international specialists gathered there, it became clear
that there was a lot of doubt about the potential of the place and the risks about building
such an installation using vague source material. The two students were not convinced
and decided not to go for a full-fledged story, but kept a critical distance when
constructing the augmented reality. They presented it as a potential story that could be
constructed and played out in the house. Two weeks after finishing their assignment an
article was published in the heritage journal OKV that was picked up by the Belgian
press. In the article the link between Pieter Bruegel and the house in the Hoogstraat was
rejected and new sources located the place of residence in another street in Brussels.
During the exam and subsequent discussions, a new dimension was explored. Morien
and Joke got hold of a study made by heritage students in 2009–2010, Pieter-Jan De
Vos, Koen Aertgeerts, Audrey Dussaard and Sandrine Herinckx. They made a study to
develop the attic of the college at the site into a new heritage attraction. Their work “The
Sint-Jan Berchmanscollege and it’s Urban Environment in Brussels” offers a lot of
possibilities. Their drawings and suggestions are interesting material to present to people
looking at the (for outsiders) inaccessible building and huge roof, imagining what is in
there and what could be in there. A 360 photo and a virtual visit of several of the potential
developments could be offered. In the exam that turned into longue conversations, we
concluded that several of the locations had similar stories to tell: not realized but poten‐
tial heritage attractions.
Just like the other students, Joke and Morien provided several pages of feedback on
what they liked and what they regretted, in particular voicing the request for better text
processing facilities. This and other feedback was used in the summer of 2016 to improve
the system, and to add a performant text editor to the app.
5 Conclusion
Even in – or precisely because of – the experimental phase of the app, inviting last year
master students in heritage studies to use this device has proven to be interesting. It is
empowering to be able to realize something that works in real life and that is more than only
a paper. The students experienced that the same archival and heuristic work they are used
to do can have a huge added value, that accompanying texts can be made and provided as
PDF. They understood the power of mobilizing their skills with creativity, but also the
importance of taking into account the role of end-users, readers and the public. Because the
app is actually intended to be used, they also had to exercise their mediating skills and
propose a marketing strategy. In the forum course of 2016–2017 the experiment will be
repeated. The challenge we will use then is to also work with processes of assessing and
formulating a statement of significance. At the end of the course, they will have experi‐
mented with two techniques that will be on the agenda in most heritage institutions in Flan‐
ders the next five years, augmented reality and frameworks of significance (Figs. 2 and 3).
ErfgoedApp 857
Fig. 2. QR code used to synchronize the panorama photo
Fig. 3. Screenshot of basic structure developed by Schroyen and Vanderschoot

References
1. Wiercx, B.: De ErfgoedApp, Innovative met iBeacons & Augmented Reality, Faro. Tijdschrift
over cultureel erfgoed, 8, nr. 3, pp. 37–39 (2015). www.erfgoedapp.be
2. Wiercx, B.: De Erfgoedapp DIY, Brussel, FARO (2015). https://s3-eu-west-1.amazonaws.
com/faronet/erfgoedapp/ErfgoedappDIY.pdf
3. Jacobs, M.: Brokerage, addressing boundaries and the new paradigm of safeguarding intangible
cultural heritage. Folklore studies, transdisciplinary perspectives and UNESCO. Volkskunde
115, 265–291 (2014)
4. Jacobs, M.: Development brokerage, anthropology and public action. Local empowerment,
international cooperation and aid: safeguarding of intangible cultural heritage. Volkskunde
115, 299–318 (2014)
5. Begeleidende tekst bij opdracht Erfgoedapp Samenwerking: Morien Schroyen – Joke
Vanderschoot. Brussel, VUB (2016)
6. Van Meerten, M.A.O.: Buiten-sporig Brussel. 50 jaar Noord-Zuidverbinding, Tielt, Lannoo
(2002)
Contextualizing 3D Cultural Heritage
James Lloyd(&)
Ancient History Encyclopedia, Horsham, UK

[email protected]
Abstract. An increasing number of cultural heritage 3D models are being made

public via the 3D-party platform, ‘Sketchfab’. This is a hugely popular way to
share cultural heritage with a wide audience. The British Museum’s model of the
Granite head of Amenemhat III has been viewed online 61,500 times and
downloaded 3,000 times (as of writing). This paper will explore Ancient History
Encyclopedia’s project to include 3D models on their website, and how doing so
helps contextualize an object, creating a deeper learning experience for the
reader of our content and the viewer of a model.
Keywords: Sketchfab 3D model Ancient History Encyclopedia

Contextualization Digital pedagogy
1 Introduction
An increasing number of cultural heritage institutions are creating 3D models and

uploading them to Sketchfab, “the leading platform for publishing and sharing 3D
cultural content online” [1] and Sketchfab actively encourages museums and cultural
heritage institutes to use its service to host their 3D models by providing them with a free
business account, enabling to make the most of the service that Sketchfab offers. Because
this field of cultural publication (open access 3D models) is still quite new though, there
are many pedagogical issues that need to be thought-through. What is being achieved,
and what do museums want to achieve, by uploading 3D cultural heritage?
This paper will focus on just one of these issues: context. A museum gallery is a
space that seeks to enhance the understanding of the objects included in it. This could
be achieved by placing objects in some form of chronological, geographical, or other
thematic order, but many more creative approaches exist. In sum, the space of the
museum gallery is a vital factor in how we engage with cultural heritage [2]. Despite
this, when we make a 3D model of a heritage object, its wider context can be over-
looked by a viewer as it is removed from the spatial context provided by a museum. As
such, this paper will focus on a questionnaire run by Ancient History Encyclopedia [3],
in order to better understand some of the issues users have with contextualizing 3D
models of cultural heritage as they appear on Sketchfab.
2 Why Sketchfab?
One of the reasons why Sketchfab is such a popular interface for cultural heritage
institutions to use (in addition to those reasons already mentioned) is its ease of use.
The website supports over 30 file types, and provides exporters for 49 different

DOI: 10.1007/978-3-319-48496-9_69
860 J. Lloyd
programmes [4, 5]. When a model is uploaded, a number of 3D settings can easily be
adjusted, such as camera field of view and lighting. Further, users are able to add
annotations to their models, enabling cultural heritage institutions to create ‘digital
museum labels’. In addition to this ease of use, studies such as that undertaken by
Adriana D’Alba and Greg Jones conclude that “using a 3D virtual environment can
create a positive impact in visitors before, during, and after a visit, and it can be a
viable alternative for those who cannot attend and learn from museums exhibitions
around the world” [6]. By categorizing and tagging their models, this virtual envi-
ronment is also searchable, but, so far, but is not as advanced as the system of 3D
Semantic Annotations as outlined by Chih-Hao Yu el al. [7]. Nevertheless, all of these
factors seem to make Sketchfab the interface of choice for open access dissemination of
3D models of cultural heritage and are the reasons why models found on this website
were used for the questionnaire.
However, there are some limitations. In order for the 3D models to load correctly,
modern internet browsers are required. This is because Sketchfab uses the WebGL
plugin to display 3D models online, and while this plugin is inbuilt in most current
internet browsers, it is not in older versions. As one teacher commented in response to
Question 15: “My browser and OS are not up to date (I’m using Windows XP) and I
can’t see the annotations. I’m not saying they don’t work, and I’m not saying they
should work on my operating system, but if your annotations have to be viewed by an
up-to-date machine will this work in the average High School classroom?”. The fact
that not all schools will have access to computers with internet browsers or internet
connections to facilitate a whole classroom browsing a 3D cultural heritage database is
perhaps a current limiter to the reach of 3D cultural heritage as an educational tool, but
nevertheless, I still agree with Benedetto et al. that there is “enormous potential
brought by WebGL for CH applications development” [8].
3 The Questionnaire
As mentioned above, this questionnaire was carried out by Ancient History Encyclo-
pedia, a non-profit organization whose mission is to improve history education
worldwide. We aim to do this by creating the most complete, freely accessible and
reliable history resource in the world. With over 2 million monthly readers, and over
300,000 social media followers we are most-read ancient history publication (online or
in print) in world. This large user-base makes us well equipped to conduct research into
how users are currently interacting with 3D models cultural heritage.
While the aim of this paper is to explore issues of context and contextualization
regarding the use of 3D models, the scope of the questionnaire was wider. As such this
paper will focus on the questions asked that were most relevant to this topic (Table 1
provides a full list of all questions).
Contextualizing 3D Cultural Heritage 861
Table 1. Copy of the questionniare

Question Multiple choice/ranking option
Demographic
1 How old are you?
2 Are you…? Male/female/other
3 What country do you live in?
4 Would you describe yourself as a…? Teacher/school
student/undergraduate/postgraduate/history
enthusiast/other
5 What is your highest qualification? High school/BA/MA/PhD/other
6 Why do you use Ancient History Personal interest/professional
Encyclopedia? interest/research for university/my teacher
asked me?
7 Which of the following topics interest you? IT/computer games/photography/cultural
heritage/history
Interacting with 3D models
8 What is your prior experience of 3D models None/aware of them/occasionally use
them/regularly use them/know how to create
3D models?
9 How easy is it to navigate this 3D model? 1–10
10 How would you rate the overall quality of 1–10
this 3D model?
11 How does using the 3D model compare to
viewing a photograph of the same object?
12 How could this 3D model be improved?
13 How easy is to navigate this 3D model? 1–10
Type of 3D model
14 How does using a 3D model of a place
compare to using a model of an object?
Using annotations
15 How useful were the annotations to
understanding this model?
Contextualizing 3D models
16 How important to understanding an object is 1–10
viewing it within its wider
historical/archaeological context?
17 How does viewing a 3D model compare to
viewing an object in a museum?
18 Have the various annotations provided
enough information about the model to
understand its wider historical context?
19 How useful (for studies/research) do you 1–5
find these 3D models?
20 Any further comments…
862 J. Lloyd
3.1 Layout of the Questionnaire

Respondents were asked to follow links to three different models currently on
Sketchfab. The first was the British Museum’s model ‘Granite head of Amenemhat III’
[9], the second was the South Etruria Tomb Survey at the Swedish Institute in Rome’s
model of ‘Tomba della Sedia at San Giovenale’ [10], and the third was Nestor F.
Marque’s model ‘Coliseo de Roma/Roman Colosseum’ [11].
The first section of the questionnaire asked general demographic questions, the
second section enquired after users’ experience in interacting with 3D models, asking
respondents to interact with the British Museum’s model ‘Granite head of Amenemhat
III’ (Fig. 1). The third section asked users to compare the experience of using a 3D
model of a cultural heritage place, compared to using a 3D model of a cultural heritage
object. Here respondents were asked to interact with the South Etruria Tomb Survey at
the Swedish Institute in Rome’s model of ‘Tomba della Sedia at San Giovenale’.
Section four asked users to comment on the use of annotations with a 3D model. Here
respondents interacted with Nestor F. Marque’s model ‘Coliseo de Roma/Roman
Colosseum’. The final section asked users a number of questions concerning 3D
models and historical/archaeology context. Since this data was solicited (the ques-
tionnaire was advertised on www.ancient.eu and Ancient History Encyclopedia’s social
media) most respondents already had an interest in history and the answers to Question
7 show that respondents’ interests were primarily historical rather than technological.
Out of 203 respondents who answered the question, ‘Which of the following topics
interest you?’, 177 answered ‘history’; 121 ‘cultural heritage’; 80 ‘photography’; 49
‘computer games; and 35 IT/programming. This data, and all following data, was
collected between 27/8/2016 22:52 and 30/8/2016 22:38. As such, the main limitation
of this questionnaire is that it was only conducted over three days, and while it gained
enough responses over that time period, a larger dataset is always welcome.
Additionally, respondents did not have to answer every question; hence why certain
questions have a higher or lower number of answers than others. There were twenty
different questions in total (Table 1). I will now focus on the answers to those questions
which are most directly relevant to exploring issues concerning 3D models and
contextualization.
4 Data
The1 majority of users thought that annotations were very useful for understanding a
3D model (Fig. 2), and thought that an object’s wider historical and archaeological
context were important to understanding an object (Fig. 3). While those responses were
perhaps expected, less than half of the respondents thought that annotations provided
enough information all of the time (Fig. 5). This helps place in context one respon-
dent’s response to Question 20 that “…the quality of the annotations are [sic] critical
to the value [of the model].”
1
For Questions 15, 17–18, which elicited a short response from respondents, answers have been
summarised with a single word or short phrase for the purposes of representation.
Fig. 1. Screenshot of Granite head of Amenemhat III by The British Museum (CC Attribution
NonCommercial) showing Sketchfab’s interface, with one of the annotations selected.
While Sketchfab enables a level of contextualization (in that models can be cate-
gorized and tagged), respondents thought it was sometimes lacking. Some of the
answers to Question 16 included the comments that: “Good, but it would be better with
context of other objects”; “Viewing it as an individual object allows one to see its
purpose and appreciate its existence, while viewing it with other historically con-
temporary objects enables one to get a general idea of what was going on at the time
and appreciate that period as a whole”; “It works on its own, but it would be a richer
educational experience if the object was given some historical context.”
However, responses were mixed, “It’s still very cool. A small amount of historical
context would be nice but comparative items would be unnecessary”; “Pay more detail
to the one and not overwhelmed by all the other artefacts.”
The responses to Question 17 are not surprising, with the majority of respondents
choosing to highlight that ‘ease of access’ was a major positive difference between
conventional museum visits and viewing 3D models of cultural heritage online (both in
terms of the physical freedom of viewing such models, and that they can be viewed in
360o, unlike the majority of objects on display in museums).
Further, it seems that respondents interpreted ‘context’ and ‘contextualization’ in a
variety of ways. Out of forty-eight responses to Question 20, one respondent com-
mented: “I would have liked for instance to know the situation of the tomb in the
landscape or to be able to situate the Colosseum as opposed to the Forum or the
Egyptian head in its place of origin, if possible?”. Sadly, visually contextualizing 3D
cultural heritage within its wider archaeological landscape is at the moment a finan-
cially unfeasible option for us (Figs. 4 and 6).
864 J. Lloyd
Fig. 2. Question 15 (146 responses)
5 Reflections on the Data
As such, it seems that the majority of respondents wanted to understand these 3D

models within their wider historical and archaeological context, but that there was
minimal desire for this to be done visually. In order to achieve this we decided to
enable users to contribute 3D models from Sketchfab to Ancient History Encyclopedia.
This means that 3D models can be embedded in our definitions and articles and be
viewed within their wider historical and archaeological context. For example, a 3D
model of a Cypriot Poppy flask created by the Ure Museum of Greek Archaeology [12]
can be embedded in a definition of Cyprus under the subheading of Bronze Age Cyprus
(Fig. 7). Users can then interact with the model while reading about its wider context,
while the model’s annotations can be used to access object specific details.

866 J. Lloyd
Fig. 7. Screenshot from Ancient History Encyclopedia’s dev server, showing the Ure Museum
of Greek Archaeology’s ‘Cypriot Poppy Flask’ model embedded in the ‘Cyprus’ definition.
Further, when users add a 3D model to Ancient History Encyclopedia they are
required to add metadata. This involves categorizing the model (such as ‘Art &
Architecture’, ‘Places’, and ‘States & Cultures’) and adding tags to it which relate to
the object itself, but also to other content on Ancient History Encyclopedia, and content
is categorized into a rough world region. This process is reviewed and ultimately
controlled by our editors in order to maintain consistency and accuracy. As such, our
search feature enables users to filter their search criteria based on these factors. This
enables the 3D models to be viewed in relation to other articles, images, videos that are
contextually related. Additionally, utilizing this metadata, an object automatically
appears as ‘related content’ on any page with which it shares tags. However, one of
Ancient History Encyclopedia’s limitations, in that it is a virtual platform, is that we do
not have access to original cultural heritage objects in the same way as museums
uploading 3D models to Sketchfab, the above process is a way which we can further
contextualize the 3D models of others.
While this is in many ways similar to how the Europeana database organizes and
presents its 3D models of cultural heritage [13, 14] we include 3D models with a
written context - an obvious differentiation between a database and an encyclopedia.
A 3D model of the Roman goddess Cybele, as presented in the Europeana database
[15] provides the viewer with plenty of information about the object (dimension,
provenance, chronology, and a certain amount of textual contextualization via a short

description of the object), but displays it very differently to Ancient History Ency-
clopedia. If that model were to be uploaded to Ancient History Encyclopedia, it would
be embedded in an article about Cybele or Roman religion, enabling a contextual-
ization of the object that would enable viewers to better understand the importance of
the object to a Roman worshipper, to give just one example. Further, there is a certain
immediacy of integration provided by WebGL and the Sketchfab model.
In the future, we plan to further contextualize our content by date and
geography/location. We also intend to link our content with the Pelagios [16, 17] and
(possibly also) Period0 databases [18], however, such plans are in their early stages. It
is hoped that the research drawn from this questionnaire will help to inform the process.
By embedding 3D models within definitions or articles, we change the focus from
solely the object to understanding how it relates to a wider body of historical and
archaeological material, but because 3D models are such a visually and interactively
stimulating medium, they additionally provide a greater understanding of the object
than a conventional 2D photograph would.
This is Ancient History Encyclopedia’s response to contextualizing 3D models, and
it will be reviewed with a follow-up survey in order to assess how successfully it has
achieved its goals, and how it might still be further improved. As such, I hope that this
paper might prompt further discussion as to the pedagogy of 3D models in cultural
heritage.
Acknowledgements. Thanks to Jason Moser for developing the integration of Sketchfab

models into Ancient History Encyclopedia, and Jan van der Crabben and James Weiner. Many
thanks to the anonymous reviewers for their constructive suggestions and corrections, any
mistakes of course remain the author’s own. Thanks too to all the cultural heritage institutes who
upload their models to Sketchfab.
References
1. https://sketchfab.com/museums. Accessed 30 Aug 2016
2. Hillier, B., Tzortzi, K.: Space syntax: the language of museum space. In: Macdonald, S. (ed.)
A Companion to Museum Studies, pp. 282–301. Blackwell, London (2006)
3. https://www.ancient.eu
4. https://help.sketchfab.com/hc/en-us/articles/202508396-3D-File-Formats?utm_source=
website&utm_campaign=upload_hints. Accessed 30 Aug 2016
5. https://sketchfab.com/exporters. Accessed 30 Aug 2016
6. D’Alba, A.: Jones, G: Analyzing the effects of a 3D online virtual museum in visitors’
discourse, attitudes, preferences, and knowledge acquisition. In: Nettleton, K., Lennex, L.
(eds.) Cases on 3D Technology Application and Integration in Education, pp. 26–47.
Information Science Reference, Hershey (2013)
7. Yu, C.-H., Groza, T., Hunter, J.: High speed capture, retrieval and rendering of segment-based
annotations on 3D museum objects. In: Xing, C., Crestani, F., Rauber, A. (eds.) ICADL 2011.
LNCS, vol. 7008, pp. 5–15. Springer, Heidelberg (2011). doi:10.1007/978-3-642-24826-9_5
868 J. Lloyd
8. Benedetto, M., Ponchio, F., Malomo, L., Callieri, M., Dellepiane, M., Cignoni, P., Scopigno,
R.: Web and mobile visualization for cultural heritage. In: Ioannides, M., Quak, E. (eds.) 3D
Research Challenges in Cultural Heritage. LNCS, vol. 8355, pp. 18–35. Springer,
Heidelberg (2014). doi:10.1007/978-3-662-44630-0_2
9. https://sketchfab.com/models/64d0b7662b59417986e9d693624de97a. Accessed 31 Aug
2016
10. https://sketchfab.com/models/554b3a527ccf4a869584f219b959555c. Accessed 31 Aug
2016
11. https://sketchfab.com/models/544c64b6445e4899a17350c949b7766a. Accessed 31 Aug
2016
12. https://sketchfab.com/models/1e6cdc4882d04fc5a4535ae8163c1ac6. Accessed 31 Aug 2016
13. http://www.europeana.eu/portal/en. Accessed 2 Sept 2016
14. Santos, P., Serna, S.P., Stork, A., Fellner, D.: The potential of 3D internet in the cultural
heritage domain. In: Ioannides, M., Quak, E. (eds.) 3D Research Challenges in Cultural
Heritage. LNCS, vol. 8355, pp. 1–17. Springer, Heidelberg (2014). doi:10.1007/978-3-662-
44630-0_1
15. http://www.europeana.eu/portal/en/record/2048709/object_HA_942.html. Accessed 2 Sept
2016
16. http://commons.pelagios.org/
17. Rainer, S. Elton, S., Leif, I.: Exploring Pelagios: a visual browser for geotagged datasets. In:
International Workshop on Supporting Users’ Exploration of Digital Libraries, Paphos,
Cyprus, 23–27 September 2012 (2012)
18. http://perio.do/
Crowdsourcing Cultural Heritage: From 3D Modeling
to the Engagement of Young Generations
Laura Inzerillo1 and Cettina Santagati2 ✉

( )
1
Department of Architecture, University of Palermo, Palermo, Italy
[email protected]
2
Department of Civil Engineering and Architecture, University of Catania, Catania, Italy
[email protected]
Abstract. Monitoring, digitizing and archiving museum artworks represent an

important socio-cultural accomplishment and an overcoming in digital preserva‐
tion today. Cultural heritage is constantly under threat of terrorist attacks and
natural disaster. The high costs related to documentation task have prevented a
constantly and massive survey activity. The low cost 3D image based acquisition
and elaboration techniques of an object, allow to carry out a 3D photorealistic
model in a short time. Therefore, a lot of museum adopted these techniques for
the artworks archiving. Crowdsourcing activities can significantly speed up
survey and elaboration procedures. If, on the one hand, these initiatives can have
a positive impact, on the other hand involve the online user with a marginal role.
In this paper we demonstrate how it is appropriate thinking the museum visitor
as “museum operator/maker” of the digital model overstepping the outcomes
achieved so far.
Keywords: Cultural heritage · 3D modeling · Structure from Motion (SfM) ·

Museum collections · Crowdsourcing
1 Introduction
Dissemination, conservation and knowledge of cultural heritage are essential elements

in its life cycle analysis (LCA), whether it is a museum, archaeological, architectural or
urban site. The possibility to schedule the updates on the state of conservation of an
artwork, ensures its own preservation for future generations. We could bear in mind
cataclysms, earthquakes, terrorist attacks, etc. that endanger the life of a physical reality
like a museum and its artworks.
The virtualization of an artwork guarantees the digital preservation and its passing
down to future generations. That is, history, identity and culture of a community are
strictly related to their cultural roots and signs of the past.
The high costs due to 3D acquisition of an artwork whether it is a sculpture, an
archaeological site, an architectural element, have not, so far, allowed a cyclical and
scheduled monitoring to guarantee a reliable analysis of the work itself. Several attempts
have been put in place to support the delicate and utterly titanic undertaking aimed at
3D archiving and reconstruction of museum collections, e.g. 3D icons project, the latest

DOI: 10.1007/978-3-319-48496-9_70
870 L. Inzerillo and C. Santagati
Uffizi museum in Florence 3D digitization ect. [1–3]. On the other hand, today there are
praiseworthy initiatives seeking in different ways to engage the online user to a museum
site, assigning him tasks to help the museum itself in some important digital goals
(Egyptian Museum of Turin, Heritagetogether project). However, the online user often
has a role still marginal and it is a small piece in the enormous reconstruction mechanism.
Our research tends to overrun the goals achieved so far. The source of cooperation is
identified in the museum or site visitor: he becomes the creator of the 3D content, acting
as a digital maker rather than a mere spectator.
Building on the results obtained from the state of the art we will demonstrate that
through the use of existing low cost 3D acquisition techniques (very easy to use and
available on the market) it is possible to create an organized system of a production cycle
that starts from the museum, involves the visitor, returns to the museum and moves to the
community.
2 The State of the Art
The last years have seen several technological innovations in the field of photogram‐
metry and computer vision techniques for the creation of useful and accurate 3D models
of objects, by combining robust algorithms, powerful computers or cloud computing
platforms. Several studies have been addressed to the exploitation of SfM algorithms
for the creation of 3D models of cultural heritage objects and sites [4–7]. The power of
Structure from Motion (SfM) techniques lies in the low learning curve for the realization
of a good dataset and in the totally automatic or semi automatic pipeline for the subse‐
quent realization of the textured 3D model. Furthermore, compact or mobile phone
cameras can also be used, with good visualization results.
One interesting project that exploits the power of SfM techniques for 3D recon‐
struction is 4D-CH-World [8]. The main goal of the project, funded under the VII Euro‐
pean framework, is to enable historians, architects, archaeologists, urban planners and
affiliated professionals to reconstruct views of historical structures starting from millions
of images floating around the web and interact with them. Furthermore, the project
foresees semantic enrichment of the end results and their subsequent export to Euro‐
peana, the European digital library, for integrated, interactive 3D visualisation within
regular web browsers.
On the other side, there is an increasingly attention in the power of crowdsourcing
for Cultural Heritage: volunteers called to absolve simple tasks that cannot be carried
out automatically by a computer and are time consuming for museums employees.
However, dealing with cultural heritage, we could consider this a niche crowdsourcing,
due to the cultural skills required to the volunteers [9].
One meaningful projects directly managed by a museum is Micro-Pasts project [10].
The project is co-led by UCL and the British Museum, since may 2016 it involves also
a pilot experience at the Egyptian Museum in Turin. The platform’s goal is to engage
traditional academics, archaeological societies, interested individual members to create
new, high quality archaeological and historical data about the human past. The project
foresees the engagement of volunteers on line assigning them simple tasks such as
Crowdsourcing Cultural Heritage 871
transcription and photo-masking on British Prehistory and British Museum collections.

As regards photo-masking, the goal is to create masks that isolate the object, improving
the 3D reconstruction (Photoscan).
Another interesting project is Heritagetogether [11]. It uses crowd-sourced digital
photographs to produce 3D models of Neolithic and Bronze Age remains from Gwynedd
and Anglesey area. It is a collaborative Arts and Humanities Research Council project,
with academic partners at Manchester Metropolitan, Bangor and Aberystwyth univer‐
sities. The inspiration for the project was the prevalence of digital camera use in the UK,
not least on mobile phones, to crowd-source archaeological research data and to create
fully textured digital 3D models using SfM techniques. Volunteers find on the web site
tutorials and a list of available software.
If the previous projects are aimed by a cultural-social instance, Stone Bridge Plaka
project and Project Mosul answer to our empathic feeling after a catastrophic event to
react and try to reconstruct lost heritage sites in order to preserve, also if virtually, the
memory and the identity of a place.
Crowdsourcing the traditional stone bridge of Plaka project [12] was aimed to collect
images on a lost heritage with the aim to virtually reconstruct it by means of SfM algo‐
rithms. On the 1st of February 2015, the central section of the Plaka stone bridge’s unique
arch collapsed due to extreme weather conditions. Then a web site has been developed for
collecting images and videos by volunteers contributors. They were processed both with
free and open source solutions (VisualSfM) and commercial ones (Agisoft Photoscan).
As regards Project Mosul [13], it rises at the end of February 2015 after the shocking
news of IS destructions of Mosul Museum. The reaction of heritage community brought
to the attempt to digitally reconstruct lost heritage starting collecting pictures and using
photogrammetry to carry out this task. It was created a platform to manage crowd-
sourced images 3D model generation, giving citizen scientists the tools they needed to
undertake this task (tutorial, available software). However, the output is a 3D textured
model which lacks of geometric fidelity, but constitutes a valuable source for visuali‐
zation, memory and documentation. Nowadays the platform has been moved to
www.rekrei.org and includes a global focus on lost or at risk heritage.
All the highlighted projects are developed on online platforms, one of the major
difficulty is the engagement of the online visitor.
However, our idea detaches from these previous experiences because directly
engages a museum visitor on site, changing and enhancing his visit experience, trans‐
forming him into a 3D cultural content creator during the visit. As will be described in
the following sections.
3 A Simple Idea for Museum Collections
The idea of identifying in the museum/archeological-site visitor, a bottom source aimed

at dissemination, enhancement and knowledge of Cultural Heritage, got leverage both
from the above described projects and the experimentation results on SfM techniques.
Although SfM 3D models have less metric accuracy and geometric precision of laser
scanning models - especially if integrated with other sophisticated techniques such as
flood-lighting instruments, X-Ray investigation, etc.-, the SfM 3D model quality

improves significantly if pictures are taken according to photogrammetric criteria,
paying attention to enlightenment conditions, and it is carried out an advanced pre-
processing before 3D reconstruction [14].
The potential of Micro-Past, Heritagetogether, Project Mosul (Rekrei) platforms, let
us to verify the practicability of a more arduous idea: to encourage the museum/site
visitor to give a personal contribution either through the creation of a data set or the
final 3D model.
Nevertheless, it was necessary to test our idea on a first target of young visitors. The
opportunity provided by the national event #invasionidigitali (thanks to which the
involved museum allows the tablet, camera, cellphone to make photos and videos)
allows us to involve our students of Palermo and Catania Engineering Schools.
The students just turned in real “digital 3D invaders”, without any cultural and tech‐
nological difficulty. The professors trained the students with the necessary technical and
critical tools on SfM techniques, so they could work on their own. The carried out models
showed expertise in the use of these techniques and capacity to investigate. Below are
showed the phases of the experimentation.
3.1 Engagement and Students/Visitors Training

At this step, thanks to the visibility provided by the national event #invasionidigitali [15]
the students felt involved in a bottom-up system of cultural heritage enhancement
through the use of social media (facebook, twitter, instagram). This has facilitated their
active participation in our project idea also thanks to the introductory seminars on
#invasionidigitali held by Dr. Elisa Bonacini, ambassador for the project in Sicily. In
addition, the simplicity and easiness in the construction of the 3D model (from the dataset
to the 3D reconstruction) encouraged the student to have a social and cultural role, being
aware of the final objective: with a little effort, the student/visitor promotes architectural
heritage and co-creates 3D cultural content of great value [16].
3.1.1 #Digitalinvasions
#invasionidigitali (#digitalinvasions) is an Italian bottom up project started in 2013 [15].
This initiative was a reaction to the cancellation, for economic reasons, of the “Culture
Week” event that allowed for free entrance to museums. Each ‘invasion’ is a mini-socio
digital event in itself with its own poster, Facebook event and hashtag. Social and digital
communication are the key to the invasions: ‘invaders’ are bloggers, amateur archae‐
ologists artists, photographers, Instagrammers, communication experts, but also
common people with a wide range of backgrounds, all with the same desire to promote
their cultural heritage through social media. By joining the Manifesto (www.invasioni‐
digitali.it), people decided to support Italy’s cultural institutions by “invading” them
with cameras, smartphones and tablets and share their cultural experiences through the
web and social media.
3.1.2 SfM Techniques

In our experiment, students were free to adopt one of the available suggested solutions
(Agisoft Photoscan or Recap360 Autodesk platform). If the first one needs a high
performance computer for data processing, the second one use the power of cloud
computing to carry out a semi-automatic data processing instead of considerably
slowing-down the computer.
A fundamental step is the realization of the photographic network of the object: it is
recommended that the angle between one shot and another is about 5–10° so that the
overlapping between neighboring frames is of about 70 %; furthermore the images
around the object should be taken with different rotations and different heights so as to
vary the angle of the shoot.
In Agisoft Photoscan the user can give inputs to the several steps of 3D reconstruction
(alignment, dense reconstruction, meshing, texturing) concerning the kind of recon‐
struction, the accuracy, the number of faces, etc. While using Recap 360 the user uploads
the images and waits the service for completing the task. The output is a 3D polygonal
textured model that can be exported for further usage in external programs according to
different formats (OBJ, Ply).
4 Results at Salinas Museum in Palermo and Civic Museum

of Castello Ursino in Catania
The students involved in the experimentation were from Drawing Course of the first
year of Environmental Engineering of University of Palermo and from Digital Survey
Course of the fourth and fifth year of Building Engineering-Architecture of University
of Catania. The students of both classes, consisting of about 80 students, lived with
enthusiasm and attractiveness the event.
In both cases students were divided into groups to any one of which was assigned a
work of art. The following images show the 3D models made by students, chosen by
different typologies and dimensions. They are visualized according 4 different modes:
textured, solid, x-ray, wireframe. This choice has been made with the intent to show
lack of holes, deformations, etc. and meshes face resolution.
Fig. 1. Torso of the Stagnone (Salinas Museum)

The data set have been made with mobile devices, digital compact camera and reflex
camera; the reconstruction resolution variable from low to high values; the data set are
made by 40–70 photos (Figs. 1, 2, 3, 4, 5, 6, 7 and 8).
Fig. 2. Statue of Zeus Ourios (Salinas Museum)
Fig. 3. Sarcofago in alabastro: Hasti the dead (Salinas Museum)
Fig. 4. Statue of Zeus sitting on a throne (Salinas Museum)

Fig. 5. Sepulchral statue end of XIV-beginning of XV century (Castello Ursino Civic Museum)
Fig. 6. Marble bust Agrippina Maggiore (Castello Ursino Civic Museum)
Fig. 7. Head of Arpocrate (Castello Ursino Civic Museum)
Fig. 8. Fragment of door cornice (Castello Ursino Civic Museum)

5 Discussion and Proposal
The models produced by the students are very impressive and, if you consider that some
have also been made with mobile or compact cameras, the result is excellent even under
the geometric profile.
Moreover, the heterogeneity of our target (students) in terms of age and cultural
background should not be underestimated: those of Palermo attend the course of
Drawing which is delivered at the 1st year of degree course in Environment Engineering
(18–19 years of age), those of Catania attend the course of Digital Survey which involves
students of 4th and 5th year of the degree course in building engineering-architecture
(22–23 years of age).
Some of the students involved, especially those of the 4th and 5th year have followed
up to the experience within the university courses also in their personal life experience:
we refer to students who have produced 3D models of artworks exhibited in European
museums (Louvre, Villa Borghese, etc.) visited during personal trips.
The results have given our idea strength aimed at identify a potential creator of 3D
cultural content in the young visitor; therefore, visitor, appropriately engaged, becomes
an indispensable resource for the development, dissemination, preservation and knowl‐
edge of the cultural heritage of society as a whole.
The crucial point is to identify how the results achieved from other crowdsourcing
experiences and our experience can find an operational synthesis within an existing
museum reality. Below the synthesis of the full process (acquisition, storage, valoriza‐
tion of the models) considering some technical and economic issues:
– Technical features of the objects/models: The geometric characteristics of the
object do not affect the 3D reconstruction of the model. Problems were found attrib‐
utable to the material that constitutes the object: for example glass and reflective
metals. Particular attention should be paid in the acquisition of objects inside show‐
cases. The features of the model in terms of graphics and metric accuracy are closely
related to the resolution of the digital camera. Therefore it would be desirable that
the museum could consider as initial investment the acquisition of two–three digital
cameras with the same technical specifications.
– Efforts/time required to the visitors: It is desirable that on average a dataset does
not exceed hundred images, so the time for the creation of a data set takes from 5 to
20 min. The timing of the loading and processing of the data set vary depending on
the hardware features/power of cloud computing. Generally, the range is from one
hour to five hour.
– Copyright issues: In order to guarantee the paternity of the creation of a model,
models will be released in CCBY license, that include in metadata the name/s of the
model creators.
– Business rationale: The business rationale has two main objectives: the first is to
realize the 3D cataloguing of the museum collection, the second is to achieve a return
in terms of number of visitors due to the media echo both of the initiative itself and
of visual/cultural/creative impact of created models.
– Operational rationale for the institutions: The digital models will be stored for the
conservation and will be disseminated as Open Data by using social media also by
video, animations, interactive navigations, etc.; each visitor will have assigned a
specific task, so there should not be any possibility of annoyance of the operation
during the acquisition process.
– Overall technical infrastructure and features of the system: The museum will
achieve the following features: digital database of photo archive of the collections
items, 3D repository enriched by metadata linked to the objects, a team of experts
that will design, check the quality of the models, maintain the technical infrastructure
and give tips to visitors. Furthermore, the museum should foresee the use of a labo‐
ratory for workstations where the visitor could upload the data set and start the elab‐
oration. And 3D prints of the model should before see.
Below we highlight some of the operational arrangements-to be implemented along
the visitor’s path so that he can be engaged and turns himself into a creator of 3D cultural
content:
– Visitor must feel involved in the role of actor from his first step at the museum
entrance: led lighting spots invite the visitor to become the 3D model author,
addressing him towards the vision of a video tutorial;
– The video tutorial, following the introductory/emotional part, is divided in three main
parts: creation of the data set, upload and configuration of software/platform param‐
eters, upload of the 3D model;
– The visitor, who participate in the digital reconstruction campaign, could get
rewarded thanks to compensation package as free tickets to the museum, score
recognition, etc.;
To ensure that the integrated system can ensure an effective response from the visitor,
it would be appropriate that the museum activates a promotional campaign through
social media and involving school students too.
6 Conclusion
The integrated system structured as above, becomes a potential enrichment not only for
the museum and the visitor/maker but also for the community and future generations.
The museum experiences a continuous updating of its data fed from a cultural, techno‐
logical and cognitive ferment bottom-driven, according to the ideals of participatory
museum. The visitor/3D maker is involved in the enrichment of the digital cultural
heritage with a starring role in a continuous upgrade of the 3D acquisition techniques,
until then unknown to him, thus contributing to the growth of knowledge.
The community and future generations will benefit from the outcomes of such large
archive, knowledge, dissemination and exploitation of artworks whose 3D visualization
may solve museums remote reachability problems or, in the future, collections digital
preservation. The operating methodological model described so far could also be
extended to the works that are conserved in warehouses stock giving them the possibility
to be disseminated and studied; to museum environments; to architectural and civil

infrastructure assets, and more generally to cultural heritage.
Acknowledgements. Authors would like to thank Marianna Marcucci, Fabrizio Todisco and
Elisa Bonacini, Digital Invasions project; Giovanna Spadafora and Sandro Garrubbo, Salinas
Museums in Palermo; Valentina Noto, Museo Civico Castello Ursino in Catania.
References
1. http://3dicons-project.eu/eng/News
2. Guidi, G., Barsanti Gonizzi, S., Micoli, L.L., Russo, M.: Massive 3D digitization of museum
contents. In: Toniolo, L., Boriani, M., Guidi, G. (eds.) Built Heritage: Monitoring
Conservation Management, pp. 335–346. Springer International Publishing, Cham (2015)
3. Maschner, H., Schou, C.D.: Virtualization and the democratization of science: 3D
technologies revolutionize museum research and access. In: 2013 Digital Heritage
International Congress, pp. 265–271 (2013)
4. Inzerillo, L., Santagati, C.: Using dense stereo matching techniques in survey. Disegnare Idee
Immagini. 47, 82–91 (2013)
5. Lerma, J., Navarro, S., Cabrelles, M., Villaverde, V.: Terrestrial laser scanning and close range
photogrammetry for 3D archaeological documentation: the Upper Palaeolithic cave of
Parpallo as a case study. J. Archaeol. Sci. 37, 499–507 (2010)
6. Lo Turco, M., Sanna, M.: Reconstructive digital modelling: the Church of Mercy in Turin.
Disegnare Idee Immagini. 41, 42–51 (2010)
7. Remondino, F.: Advanced 3d recording techniques for the digital documentation and
conservation of heritage sites and objects. Change Over Time 1(2), 198–214 (2011)
8. Ioannides, M., Hadjiprocopis, A., Doulamis, N., Doulamis, A., Protopapadakis, E.,
Makantasis, K., Santos, P., Fellner, D., Stork, A., Balet, O., Julien, M., Weinlinger, G.,
Johnson, P.S., Klein, M., Fritsch, D.: Online 4D reconstruction using multi-images. ISPRS
Annals of Photogrammetry Remote Sens. Spat. Inf. Sci. II-5/W1(1), 169–174 (2013).
Available Under Open Access
9. Biella, D., Sacher, D., Weyers, B., Luther, W., Baloian, N., Schreck, T.: Crowdsourcing and
knowledge co-creation in virtual museums. In: Baloian, N., Zorian, Y., Taslakian, P.,
Shoukouryan, S. (eds.) CRIWG 2015. LNCS, vol. 9334, pp. 1–18. Springer, Heidelberg
(2015). doi:10.1007/978-3-319-22747-4_1
10. Bonacchi, C., Bevan, A., Pett, D., Keinan-Schoonbaert, A., Sparks, R., Wexler, J., Wilkin,
N.: Crowd-sourced archaeological research: the micropasts project. Archaeol. Int. 17, 61–68
(2014)
11. Griffiths, S., Edwards, B., Karl, R., Labrosse, F., Miles, H., Moeller, K., Roberts, J., Tiddeman,
B., Wilson, A.: Crowd-sourcing archaeological research: HeritageTogether digital public
archaeology in practice. Internet Archaeol. 40 (2015). http://dx.doi.org/10.11141/ia.40.7.3
12. Stathopoulou, E.K., Georgopoulos, A., Panagiotopoulos, G., Kaliampakos, D.:
Crowdsourcing lost cultural heritage. ISPRS Annals of Photogrammetry Remote Sens. Spat.
Inf. Sci. II-5/W3, 295–300 (2015)
13. Vincent, M.L., Flores Gutierrez, M., Coughenour, C., Lopez-Menchero Bendicho, V.M.,
Remondino, F., Fritsch, D.: Crowd-Sourcing the 3D digital reconstructions of lost cultural
heritage. In: 2015 Digital Heritage, vol. 1, pp. 171–172 (2015)
14. Gaiani, M., Remondino, F., Apollonio, F.I., Ballabeni, A.: An advanced pre-processing
pipeline to improve automated photogrammetric reconstructions of architectural scenes. J.
Remote Sens. 8(3), 178 (2016)
15. Bonacini, E., Marcucci, M., Todisco, F.: #DIGITALINVASIONS. A bottom-up crowd
example of cultural value co-creation. In: Proceedings of the First EAGLE International
Conference 2014 on Information Technologies for Epigraphy and Digital Cultural Heritage,
29–30 September and 1 October 2014, Paris, pp. 265–284. Sapienza Università Editrice
(2014)
16. Bonacini, E., Inzerillo, L., Marcucci, M., Santagati, C., Todisco, F.: 3D #Digital Invasions: a
crowdsourcing project for mobile user generated content. Furnace J. 2 (2015)
Training Schools for Conservation of Cultural Heritage:
Between Expertise, Management and Education
Anna Lobovikov-Katz1 ✉ , Gumersindo Bueno Benito2, Vanesa Marcos Sánchez2,

( )
Joao Martins3, and Dalik Sojref4

1
Faculty of Architecture and Town Planning, Technion - Israel Institute of Technology,
Technion City, 32000 Haifa, Israel
[email protected]
2
Foundation Santa María la Real Historical Heritage, Aguilar de Campoo, Palencia, Spain
{gbueno,vmarcos}@santamarialareal.org
3
Faculty of Sciences and Technology (UNL) and UNINOVA, Lisbon, Portugal
[email protected]
4
WTTC, Berlin, Germany
[email protected]
Abstract. Training schools make an important feature of the European research

landscape, fostering exchange in frontier research, and building basis for further
research and development. How the cross-area, management and educational
issues can be effectively put together in training schools, and specifically, for the
benefit of a multi- and interdisciplinary field of conservation of cultural built
heritage? This paper showcases the experience from the first lessons of the COST
(European Cooperation in Science and Technology) Action i2MHB (Innovation
in Intelligent Management of Heritage Buildings), to examine and suggest tools
useful for further multidisciplinary synergies and networks in this and other
frameworks.
Keywords: Training school · Conservation of cultural built heritage · Education ·

Interdisciplinary
1 Introduction - EU Training Schools in the Frameworks

of Cost Actions
Training schools are a valuable tool in order to enrich sustainable development. They
can be used as a lifelong learning tool or as a fast track to scientific updating. Either
use is valuable, with its own singularities, but targeted to quite distinct audiences.
According to UNESCO, “Every person, at every stage of their life should have life‐
long learning opportunities to acquire the knowledge and skills they need to fulfill
their aspirations and contribute to their societies”. The UNESCO’s new education
agenda strategic objectives for 2014–2021 encompass three major strategic objec‐
tives: (1) Developing education systems to foster quality and inclusive lifelong
learning for all; (2) Empowering learners to be creative and responsible global citi‐
zens and (3) Shaping the future education agenda [1]. On the other hand, rather than

DOI: 10.1007/978-3-319-48496-9_71
Training Schools for Conservation of Cultural Heritage 881
providing general training or education, training schools should provide intensive

training in emerging research topics.
The issue of education and training, along with youth and sport, is addressed in the
European Commission through the Erasmus+Programme [2]. Europe recognizes that
providing citizens with the necessary skills will allow them to play an active society role
and achieve personal fulfillment. The Erasmus+Programme has three major key actions:
(1) Mobility of individuals; (2) Cooperation for innovation and the exchange of good
practices and (3) Support for policy reform. Regarding training schools key action 1 is
the most interesting vector; thus it considers mobility of learners and staff, Erasmus
Mundus Joint Master Degrees and Erasmus+Master Loans. Regarding the mobility of
learners and staff a Vocational Education and Training (so called VET) traineeship
abroad is considered up to a maximum of 12 months. These vocational training schools
are set for both apprentices and students. It is mandatory that the vocational training
takes place in another country, where trainees are hosted at a workplace (in an enterprise
or other relevant organization) or at a VET school (with periods of work-based learning
in an enterprise or other relevant organization). Several H2020 projects offer training
schools, within their activities, dedicated to early-stage researchers including interna‐
tionally renowned scientists and leading industrials. These training schools allow young
researchers to further training and allow them to acquire new complementary skills in
order to enhance their career development (Fig. 1).
Fig. 1. COST Training Schools: the main focus of trainees-trainers knowledge and ideas’
exchange in Training Schools (TS) in Conservation of Cultural Heritage (CCH) © A. Lobovikov-
Katz
European Cooperation in Science and Technology (COST organization) is the

longest-running European framework. It has been supporting trans-national cooperation
across Europe since 1971 [3]. In its role as the breeding ground for trans-sectoral multi‐
national co-operation projects, COST delivers the best frame to set an effective platform
882 A. Lobovikov-Katz et al.
to share the knowledge on related research topics. Through COST Actions, COST
supports networking activities with particular attention to research training and
exchange. The COST Action functions to accomplish a true multidisciplinary effort and
consequently intends to overcome the fragmentation within dedicated scientific and
technological research field and thus deliver to the young members of the Action a new
horizon to nourish innovative flagship initiatives in an international context. COST
Actions have several networking tools at their service:
• Conferences
• Workshops
• Dissemination activities
• Short-term scientific missions
These networking tools let researchers improve their ideas by sharing them with their
colleagues, providing valuable inputs for their research, career and innovation. The
COST Association view of a training school is an instrument that provides intensive
training in emerging research topics within the laboratories and organizations involved
in a particular COST Action. Their participants are mainly, but not exclusively, young
researchers involved in COST Actions. However, training Schools can also cover
appropriate re-training as part of life-long learning. In the year of 2014, COST has
financed 370 Actions and 367 training schools have been organized [4].
2 Content and Management Outline of the First

i2MHB Training School
Currently composed of 36 countries, the COST Association has created a multidisci‐

plinary team of researchers, experts, companies and public administrations of 23 coun‐
tries linked to various aspects of heritage management, aimed at implementing actions
to more intelligently manage heritage. COST ACTION TD1406 Innovation in Intelli‐
gent Management of Heritage Buildings Training School took place in Aguilar de
Campoo, Palencia (Spain) on January 27th to 29th 2016. This Training School was based
on work from COST Action TD1406, supported by COST (European Cooperation in
Science and Technology).
COST (European Cooperation in Science and Technology) is a Pan-European, inter‐
governmental framework whose mission is enabling breakthrough scientific and tech‐
nological developments leading to new concepts and products, thereby contributing to
the strengthening of Europe’s research and innovation capacities. Foundation Santa
María la Real Historical Heritage is strongly interested in creating awareness of the
current challenges of the heritage sector, and here is where this kind of Training School
has its origins. Foundation Santa María la Real Historical Heritage, a Project from
Castile and León, emerged in 2014 as the result of the union between two reference
entities with extensive experience in the field of heritage: Foundation Santa María la
Real Historical Heritage. While maintaining the character of its predecessors, the new
institution provides services at the national and international level, with its main focus
on the study, restoration, conservation and promotion of social, natural and cultural
heritage. The combined efforts of the two participating institutions bring together a wide
range of interests and ongoing or past activities, of which more than 500 involve the
restoration and enhancement of heritage assets in the nine provinces of Castile and León,
as well as at the national level, in different regions.
Triggered by the opening of the COST ACTION call, the necessary efforts for the
organization of the activities that would take place during three days were made, and
the request to the involved experts for their collaboration was carried out. The next step
was the selection of the most suitable trainees, which was accomplished by a board of
COST ACTION members. The applicants were carefully selected, on the basis of the
motivation letter and CV that they had previously submitted.
English was the second language in the headquarters of Foundation Santa María la
Real Historical Heritage during the Training School. 13 students and doctoral students
from Spain, and 13 additional ones from different European countries were awarded a
scholarship by the European COST association to participate in a training school on
innovation in the field of heritage. The Training School enabled the learning of new
techniques, methods and advances in heritage management, while working to generate
research areas and create networks for specialized contacts.
During the first working day, the participants were able to get close to the Foundation
experience in the field of heritage management, as the entity director, Juan Carlos Prieto,
explained them how the entity has worked “for more than thirty years on the study,
restoration, conservation and dissemination of heritage”. They were told that later,
owing to the addition of other entities and administrations, the entity acquired new skills,
and evolved from restoration to preventive restoration and intelligent heritage manage‐
ment. Projects like Monitoring Heritage System (MHS), installed with success in about
fifty buildings, enabled the transition. With regards to MHS, not only were the trainees
told how the system works, but Dr. M. Chiriac, the engineer of Foundation Santa María
la Real Historical Heritage, explained them how the system was developed, the way the
sensors act, and their application in heritage building monitoring.
On the second day, the students had the opportunity to see in situ how the system
works: they visited MHSLab, an intelligent lab created by Foundation Santa María la
Real Historical Heritage, which was established in the Romanesque chapel of Canduela.
The objective of MHSLab, as the trainees were told, was to test the effectiveness of
different types of techniques and materials used in the field of heritage. They also had
the opportunity to get to know the Iberian Peninsula Romanesque Encyclopaedia expe‐
rience, a true cultural project where more than a thousand researchers have collaborated
to catalogue many of the Romanesque remains in the Peninsula. The Encyclopaedya
can be found in international bookstores and libraries.
Pedro Martín (OMG), a researcher at Cartif Technology Center, gave a talk of his
work on 3D documentation of historic buildings. Marija Damjanović, from Croatia,
focused her presentation on the transformation of historic landscapes; and Laura Daví,
from the University of Barcelona, introduced her research focus on the monastery of
San Cugat del Vallés.
Later that day, the Exhibition Centre ROM was visited, and the Israeli architect Anna
Lobovikov-Katz, who had developed a new method to improve heritage conservation
analysis through technical methods of drawing, participated as a speaker.
On the third day, trainees learned from the following interventions:

The impact of cultural infrastructures. The case of the Human Evolution Museum
in Burgos. Alejandro Sarmiento. Manager of Atapuerca System and HE Museum.
A conference on a Digital Information Model for the Knowledge and Management
of Immovable Cultural Heritage. Patricia Ferreira Lopes. Architect. Phd student at
Higher Technical School of Architecture of Seville. Research group HUM799 “Strat‐
egies for the heritage knowledge”. A talk on Tourism and Heritage: Building Smart
Destinations. Juan Vicente Calle Lamelas. Phd student in Tourism. University Rey Juan
Carlos Madrid.
The results of the previous day’s practical, conclusions and demonstration of appli‐
cation of Rapid Learning Methodology for Conservation (Preservation) of Cultural
Heritage – Understanding through Drawing. Dr. Anna Lobovikov-Katz, Arch., PhD,
Technion, Israel.
And finally we ended with a round table discussion on Conclusions of the training
school: Challenges for the Early Career Investigators (ECI). During those three days,
different topics were addressed: for example, the methods developed by technicians from
the University of Belgrade to counteract the effects of earthquakes on heritage, or to
fight biological problems affecting conservation. In addition, some speakers explained
practical cases, such as the management of Monsalud monastery and the Roman site of
Ercávica in Cuenca.
The quality of the Training School was assessed by means of an anonymous survey
to the trainees. General results of the survey are shown in Figs. 2, 3, 4 and 5.
Fig. 2. Q: Did the content and activities carried out in the training school correspond to your
expectations?
Fig. 3. Q: Did the trainers explain in a clear and concise way the activities in each section?
Fig. 4. Q: Were the professionals responsible for the training school close to the students,
fostering a good work ambience and participation?
Fig. 5. Q: Were the installations’ conditions adequate for the seminar?
Both trainers and trainees of this Training School represent wide range of the areas of
knowledge in research and practice within the field of conservation of cultural heritage
(CCH).
3 Interdisciplinary Research and Education in Conservation

of Cultural Heritage - The Interdisciplinary “Language”
for Training Schools (TS)
This section of the paper examines the selected aspects of TS in CCH and their educa‐
tional value, keeping the following issues in mind:
• Multi- and interdisciplinary character of CCH and CCH training schools
• Main lines of inter-area contributions of education value within the CCH complex
• CCH training schools and contemporary education.
As mentioned in the Introduction, training school is an instrument that provides intensive
training in emerging research topics within the laboratories and organizations involved
in a particular COST Action. This section focuses on the trainee as the main beneficiary
of a Training School, through direct instruction. Analysis of the leading role of COST
Action management and of local organizers of TS is beyond the scope of this paper,
while trainers-related issues are mentioned here merely in the context of trainee-centered
analysis.
3.1 Multi- and Interdisciplinary Training in CCH Training Schools

Among many interdisciplinary areas of human activity, the modern field of conservation
of cultural heritage holds a unique position, comprising literally the entire spectrum of
STEM, SSH and the arts. Training schools in CCH reflect the multi- and interdiscipli‐
nary character of this field Therefore, many organizations involved in a single CCH COST
action, usually represent a wide spectrum of STEM and SSH fields, thus adding their
input to the interdisciplinary structure and content of COST training schools in general,
including those in the framework of COST ACTION TD1406 Innovation in Intelligent
Management of Heritage Buildings (i2MHB). Analysis of the COST i2MHB training
schools showed a wide range of CCH areas of expertise of both trainers and trainees,
along with multi- and interdisciplinary cross-area topics introduced in the school program.
3.2 The Audience
Actors. With a view to contribution to CH conservation through educational activities,

several main types of actors and their inter-area contribution can be defined. The
following main groups of actors can be delineated:
• CCH researchers
• CCH practitioners
• CCH students
• Non-CCH public of diverse background.
Conservation Community. In conservation of CH, research is an integral part of resto‐

ration practice, and at the same time, contribution of education to research in CCH has
been examined in several publications [5]. With regard to actors from the conservation
community (CCH researchers, CCH practitioners and CCH students), we come up with
the following scheme (Fig. 6). It presents a multi-beneficial contribution between and
within different departments of CCH-activity (practice, research, education) and CCH-
actors (practitioners, researchers, students).
Some inter-area contributions are well-explored phenomena, e.g. a dually-beneficial
CCH students-CCH practice exchange. Experience in exposure of learners to actual
conservation process in curricula or courses in conservation of cultural heritage, has
shown good results [6, 7]. Thus, students contribute to actual conservation of cultural
heritage, and at the same time enrich their knowledge and understanding in a way which
could be hardly achieved in the classroom, through e-learning. In training schools,
combined participation of researchers and practitioners, both on the side of lecturers and
trainees, reinforces such exposure to the diverse areas of activities. In long-term educa‐
tion, even in a usual semester course, it contributes to the development of cross-area
abilities, which might be of benefit to the main area of interest of a trainee, e.g. a conser‐
vation practitioner, exposed to a research project, could develop typical research qual‐
ities. According to the COST rules, the duration of a training school ranges between
3 days to one week. This short duration of training schools adds to the intensity of inter-
and intra-area exchange between diverse groups of actors involved.
Fig. 6. “Areas-Actors-Activities” scheme: practice-research-education Interaction in

Conservation of Cultural Heritage (CCH) © A. Lobovikov-Katz
Non-conservation Public of Diverse Backgrounds. With regard to multiple types of

actors in CCH complex, another group should not be overlooked: the non-conservation
public. Generally, its members can be subdivided into several sub-groups including (a)
general public; students of different levels of education, especially (b) secondary and
(c) tertiary levels; (d) non-conservation researchers. The groups “c” and “d” are of
particular interest, especially because of inter- and multidisciplinary nature of conser‐
vation of cultural heritage. Involvement of sub-groups “c” and “d” (university students
and non-conservation researchers) might open new horizons for emerging researchers
and engineers, and also contribute to their understanding of interdisciplinary cooperation
in the modern reality. CCH learning contributes to understanding of interdisciplinary
cooperation in all areas, also those non-related to conservation, among both “c” and “d”
groups. Inclusion of non-conservation researchers in the field of conservation of cultural
heritage is not a new phenomenon. Their active participation in training schools allows
their re-training as part of life-long learning.
3.3 Multiple and Diverse Roles of Trainees in Training Schools
Following good TS practices, and specifically in the course of COST Action MP1004:
“Hybrid Energy Storage Devices and Systems for Mobile and Stationary Applications”,
2011–2015, different types and levels of involvement of trainees have been encouraged
in the COST Training School in Aguilar de Campoo.
In addition to their usual learning role, young participants can be encouraged to
propose topics for lectures and for round tables sessions which might be of interest,
chairing the round tables discussions (community of young researchers as a breeding
ground for new ideas and approaches). They can be also encouraged to suggest the
trainers or “invited speakers” etc. In the COST Training School in Aguilar de Campoo
trainees were encouraged to give talks/lectures along with their trainers. These
correspond to one of the main COST ideas - to pave the way for active involvement of
young researchers.
Figure 7 outlines multiple optional roles of trainees (versus trainers) in a Training
School (TS) which might include their involvement in the planning stage before TS (a),
active participation in the TS activities (b), and evaluation of results after the completion
of TS (c). Exposure of trainees to planning, and undertaking activities of combined
content-management character provides excellent training ground for the contemporary
research reality. Performing such diverse roles as trainee and trainer, student and
lecturer, allows the TS trainees to widen their horizons, and to improve their learning
abilities through active learning.
Fig. 7. Encouraging multiple and diverse roles of trainees in training schools before, during and
after TS activities © A. Lobovikov-Katz
3.4 Training Schools (TS) and Education

Modern education theory and practice provide a wide range of learning approaches and
practices. Several types of learning relevant to CCH training schools can be outlined:
• Class learning activities
• Laboratory
• In-situ learning on historic or conservation sites
• E-learning
• Blended learning (combining frontal instruction with online learning activities).
Bringing TS trainees together in real environment, thus allowing their immediate
communication with other trainees and trainers, - is an important advantage of training
schools. Therefore, distance learning and e-learning - the innovative contemporary ways
of learning - are less relevant to CCH Training Schools. However, blended learning
might be helpful, and especially through the inclusion of online activities before or after
the TS period, and in non-virtual instruction which takes place on TS venue. Such
delayed-in-time linkage of online and offline learning would be a complex multi-disci‐
plinary education development. In addition, blended learning might be also instrumental
for real-time blended learning during TS period, in order to facilitate understanding
between trainees and trainers of diverse conservation backgrounds. In this case, on-line
tasks might be differentiated to suit the diverse parts of target audience, and involve
“flipped classroom” methods.
On-site learning activities are indispensible in CCH, which is in line with innovative
development in education of recent decades [8]. Pre-planned combination of “passive”
and “active” learning activities in diverse types of learning and in different learning
situations, and especially the in-situ learning in CCH research and practice held on
heritage conservation sites, can result in innovative results [9].
Training School (TS) in CCH comprises diverse components, and gives trainees a
unique access to:
• Diverse types of learning activities (class/on-site/laboratory; active/passive)
• Inter- and multidisciplinary learning material
• Local cultural heritage and its conservation
• Diverse types of knowledge (CCH research; actual CH projects)
• Immediate learning and scientific exchange with senior CCH experts
• Diverse teaching methods showcasing educational trends from different universities
and other organizations
• Combining several roles in TS (optional), e.g. learning (trainee), teaching (giving a
presentation), scientific management (contributing to a TS planning; moderating a
session)
TS goals can be easily formulated as deriving from the contemporary CCH requirements
and EU policy in research, education and training. However, the open-end character of
TS activities can complicate defining and achieving specific learning objectives. The
application of educational theories and approaches, commonly used in “regular” educa‐
tion on different levels, should be specifically adjusted to CCH TS. According to the
authors’ experience, TS trainees are often characterized by high learning motivation and
multi-disciplinary educational background. These and other data often contribute to
effective learning and creative knowledge exchange between trainees, and trainers, and
support the overall dynamic character of training schools in the conservation of cultural
heritage.
4 Conclusions
The summarizing document of COST strategic Workshop The Safeguard of Cultural

Heritage pointed out that ‘the need for trans-disciplinary cooperation and training of
young scientists to strengthen the European research arena in the field of cultural heritage
is a recurrent theme’ [10]. Ultimately interdisciplinary character of modern conservation
of CH, involving different fields of science, technology, humanities and arts, tight
connection between research and practice typical of this field, along with often multi-
and interdisciplinary profile of CCH experts, have a challenging impact on CCH
Training Schools.
Following the continuing development in conservation of cultural heritage, and educa‐
tion, there would be a lasting need for reviewing and re-developing CCH training. Training
Schools in conservation of cultural heritage can provide a unique testing ground for
improvement and development of training in interdisciplinary research and development
outside the CCH domain, and, in turn, contribute to the preservation of the world heritage
through training researchers, engineers, architects, archaeologists and conservators.
Acknowledgements. This article is based upon work from COST Action TD1406 i2MHB -
Innovation in Intelligent Management of Heritage Buildings, supported by COST (European
Cooperation in Science and Technology).
References
1. UNESCO Education Strategy: United Nations Educational, Scientific and Cultural

Organization (2014)
2. Erasmus+Programme Guide: European Commission (2016)
3. About COST: C. Halén, COST Office (2014)
4. COST Annual Report 2014: COST Association (2015)
5. Lobovikov-Katz, A.: Heritage education for heritage conservation (contribution of
educational codes to study of deterioration of natural building stone in historic monuments).
Strain Int. J. Exp. Mech. 45(5), 480–484 (2009)
6. Lobovikov-Katz, A.: The virtual and the real: e-learning in interdisciplinary education – the
case of cultural heritage. In: The 13th Annual MEITAL National Conference “New
Technologies and Their Evaluation in Online Teaching and Learning”, Technion - Israel
Institute of Technology, Haifa, June 2015 (2015). http://meital.iucc.ac.il/conf2015/papers15/
A3_3.pdf
7. Keith, D.A.: Reflections on eight semesters of employing service learning in an undergraduate
historic preservation course. Preserv. Educ. Res. 4 (2011)
8. Levin-Peled, R., Sagy, O., Lobovikov-Katz, A., Chitaiad, T.: Guidelines and tips for creating
an interesting and appealing learning process for young students. In: Educational Linkage
Approach in Cultural Heritage, ELAICH, ENPI 150583 (2010)
9. Lobovikov-Katz, A., et al.: Tangible versus intangible in e-Learning on cultural heritage: from
online learning to on-site study of historic sites. In: Ioannides, M., Magnenat-Thalmann, N.,
Fink, E., Žarnić, R., Yen, A.-Y., Quak, E. (eds.) EuroMed 2014. LNCS, vol. 8740, pp. 819–
10. Fioravanti, M., Mecca, S. (eds.): Proceedings of COST Strategic Workshop on the Safeguard
of Cultural Heritage: A Challenge From the Past for the Europe of Tomorrow, 11th–13th July
2011 Florence, Italy (Proceedings e report; 80). Firenze University Press, Firenze (2011)
Educational Creative Use and Reuse of Digital Cultural
Heritage Data for Cypriot UNESCO Monuments
Marinos Ioannides1, Pavlos Chatzigrigoriou1 ✉ , Vasilis Bokolas1,

( )
Vasiliki Nikolakopoulou , Vasilis Athanasiou , Eirini Papageorgiou1,

1 1
Georgios Leventis1, and Christian Sovis2

1
{marinos.ioannides,p.chatzigrigoriou,v.nikolakopoulou,
vasilis.athanasiou,e.papageorgiou,georgios.leventis}@cut.ac.cy,
[email protected]
2
7Reasons, Hauptplatz 11, 3462 Absdorf, Austria
[email protected]
Abstract. Nowadays, there is a rising demand of reusing the constantly enriched

information from heritage digitalization in different ways. One of the objectives
of the EU Europeana Space project is the development of a holistic approach for
educating people (grown-ups and kids) on Monuments that are listed at UNESCO
world heritage list, in Cyprus. The proposed model action is based on the cross
cultural approach which, at the same time, responds to the contemporary peda‐
gogical and methodological directions. The system uses innovative digital
heritage resources to help the user learn about the different phases of the monu‐
ment, the history, the architectural value and the conservation stage. The result is
a responsive educational platform, where every Monument is a different course
and every course is addressed to different age groups. Moreover, part of our future
work is the evaluation of the platform by particular groups of our target users.
Keywords: 3D models · Education · Digital heritage · Cross-cultural · CH data

reuse
1 Introduction
Objective of the project is the holistic approach of a listed UNESCO World Heritage
Monuments in Cyprus, for development of a platform to educate people (grown-ups and
kids). The system uses innovative digital heritage resources, in order to help the user,
in a UX friendly way, to learn about the different phases of the monument, the history,
the architectural value and the conservation stage. The resulting platform is accessible
through smart devices and desktop computers, (in the frame of “Bring Your Own
Device” a.k.a. BYOD). The main challenge was the reuse of Europeana Data (Pictures
and 3D objects) in a way that the information on the platform would be comprehensible.
Computer technologies allow various visualizations and digital re-constructions of
archaeological monuments and sites. These models can be integrated in a demonstration
as an educational user friendly platform, a tool that addresses Cultural Heritage (CH)

DOI: 10.1007/978-3-319-48496-9_72
artefacts’ reconstruction by using the Europeana (and other creative commons) collec‐
tions. The work started with a preliminary search on Europeana and other online
management galleries and the identification of objects/monument for the demonstration.
It then used the images and models that can be found on Europeana as thumbnails and
that can be reached through Europeana on the holders’ web-sites. When the object is of
high quality and ready, then it can be offered for free use to education, architecture,
historians, civil engineers, students, etc. The tool will be available by the end of the
project as a Pilot test platform for primary, secondary, tertiary educational levels, and
also to support vocational training (for archaeologists, preservation, protection of monu‐
ments and sites), and the tourist industry (for example guides). Also, an evaluation plan
has been set up in order for us to have the first results from our target users, their reactions
to the content and how the platform could be implemented as a teaching procedure.
2 Aims - Limitations
The basic aim of the project is the development of a completed system of a digital
information, education and entertainment (edutainment) for adults, adolescents and
teenagers concerning the most representative monuments and landmarks of Cyprus,
reusing existing data from Europeana.
The proposed UNESCO monuments digital learning system will further aim to:
1. Help users to learn the history of the UNESCO monument
2. Educate users in architecture and cultural heritage
3. Raise awareness on protecting and conserving our heritage
4. Promote the UNESCO monuments
5. Inform on risks that threaten the monuments
For the development of the Educational Tool, we are re-using Digital Heritage Data
(DHD), starting from simple pictures of the Monuments and - at the highest end - 3D
objects. A complete recording of a monument requires the first step to be a digitization
to develop the 3D model. 3D objects can be presented by plethora of methods and tech‐
nologies depending on factors that influence the monument and the suitability. Three
most common factors that need to be calculated before initializing an approach to a 3D
reconstruction are the following: (a) Complexity in size and shape, (b) Morphological
complexity (level of detail) and (c) Diversity of raw materials.
Since in this project’s aim is the re-use of CH data that thrive through Europeana
collections, we contacted an in depth research for this content in Europeana’s Reposi‐
tories. Apparently, Europeana’s content is inefficient for 3D image reconstruction.
Furthermore, the Digital Heritage Surveying Data (DHSD) is also limited as well as
several other techniques that are also helpful in the 3D reconstruction procedure of a
monument, as empirical techniques, topographic techniques and rectified photography.
Educational Creative Use and Reuse of Digital Cultural Heritage Data 893
3 Educational Approach
Taking into consideration all the aforementioned limitations, we adopted our project to
the content we had available. Certain images and 3D models of UNESCO’s monuments
in Cyprus were used in order to form the educational content of the platform. The
presented educational tool responds to the multicultural features of present era but at the
same time to the contemporary pedagogical and methodological directions.
Nowadays, the challenge of re-using digitized cultural heritage content from Open
Repositories is an issue that concerns scientists, practitioners, professionals and insti‐
tutions of the related field. One of the areas that re-use can yield fruit is education. On
the one hand, arts and cultural heritage education has been regarded as the key factors
in development of the knowledge society and creative ability [1]. On the other hand, the
ongoing development of Information and Communication Technologies (ICTs) is
providing new opportunities of representation and interpretation of cultural heritage.
Thus, learning about cultural heritage with the mediation of ICT tools and methods - in
our case the e-learning platform enriched with multimedia content - can be affective in
all educational levels by using appropriately the available ICT affordances.
Information technology and the ‘culture of the image’ gradually and steadily gain a
key role in school learning while at the same time there is a constant fervent effort in
the modern educational systems to functionally and effectively integrate the ‘new media’
both as teaching tools and as a distinct separate school subject (in ventures of a refor‐
mation/review of the content of the existing curricula). Besides all this, the most impor‐
tant reason for using digital tools in education is because it constitutes an entraining way
for the acquisition of knowledge and it can compose a refreshing change to the daily
school routine enhancing children’s enthusiasm [2]. Apart from underage learners, we
also aim to engage adolescents, adults as well as vocational trainers in learning and
becoming aware of these historic monuments. Learning about the culture, the living
conditions, the human creations and attitudes in the course of time, understanding the
historic-cultural events and the evolution of society, and finally discerning what has
survived at the sociocultural surroundings, is the quintessence of arts and cultural
heritage education. But in order to achieve the favorable learning outcome, there are
four learning approaches that we employed for ensuring the values of ICTs in cultural
heritage education: (1) Personalized, inquiry-based learning approaches (2) On-site and
anywhere learning experiences (3) Interdisciplinary learning approaches and (4) Collab‐
orative learning experiences [1].
Each one of these approaches has its origins to the main changes brought by the
massive and precise digitalization of Cultural Heritage in the recent years. Every single
cultural artifact becomes digital in a way that its primary ontological value is expanded.
For instance, images of such artifacts are no longer bi-dimensional; their detailed and
zooming abilities allow users to study and interact with them in any preferred dimen‐
sional level. What is more, zooming out of those images puts cultural artifacts into a
wider context that reveals where they lie on or have been created. These are perspectives
that can adopt innovative approaches to teaching and learning [1]. In the sections below,
it is analyzed how we manage to approximate the aforementioned approaches to the
online educational platform.
4 Methodology
First of all, it was very important that the design of the online educational platform would
follow the “bring your own device” (BYOD) [3, 4] approach, as the most common and
easier to apply method. BYOD refers to “technology models where students bring a
personally owned device to school for the purpose of learning” [5]. There has been an
increasing interest in supporting BYOD in schools by growing support from students,
parents, and teachers. A US survey conducted by Project Tomorrow, shows that over
60 % of middle and high school students prefer to use their own devices for learning
unlike school supplied laptops or netbooks [6]. Another survey by Bradford Networks
for educational institutes in US and UK [7] showed that 78 % use devices in their school
systems “for personal use by teachers and students,” while 72 % stated that students use
the devices to complete class assignments. The results indicate great acceptance of
BYOD approach and that might be due to the very purpose of any educational sector to
provide knowledge with every means of information possible. Nevertheless, although
BYOD is generally considered to help promote better outcomes via a more personalized
learning and act as a connection between home, school and other spaces, how BYOD
can offer to inquiry-based pedagogical practices in authentic learning environment is
still immature [3].
Therefore, in order to achieve a more personalized and possible on-site and anywhere
learning experience, we developed a platform compatible to all smart devices (android,
iPhone, iPad, windows tablets). A responsive web site was decided to be used, with the
Fig. 1. Screenshot of the web site of Panagia Assinou online course. Orthophotographs,
photogrammetric pictures, 3D sections and other surveying products were used for educational
purposes.
capability to run on desktop and portable devices. The design of the platform is based on
an Educational Theme (WordPress) with CSS reconfiguration. The decision for an Educa‐
tional Theme was taken in order to use an environment familiar to most of the visitors
(both students and teachers). Since it is widely regarded that the most common used educa‐
tional platforms at the internet are Coursera, Udacity, edX etc. [8, 9], we use similar User
Interface design so that is familiar to the users. This way we cover the UX Design require‐
ment (Fig. 1). Having in mind that the challenge is to use the Europeana Data (mostly
photogrammetric pictures and 3D objects) in a way that the information on the platform
will be comprehensible by the users, data should be presented in an appropriate way so to
cover all the users’ spectrum. Most of the data have little metadata information and they
lack history and cultural value description (semantics). Even when we have this informa‐
tion, we need to adjust the data to make the information useful for educators and people
Fig. 2. A detailed site map was the first step to design the site.
that are not familiar with cultural heritage. The first step was to design a detailed Site Map,
taking under consideration the UX Design guidelines (Fig. 2).
As one can see at the demonstrator site (wp.digitalheritagelab.eu), every monument
is a different “Course”. Every course has a number of lessons, just like in any online
education portal. The information about the monument is divided to separate lessons,
as follows (Fig. 3):
1. Lesson about the History of the Monument.
2. Lesson about the Local History: short reference on the local history through the
literature study.
3. Video: short dramatized video, which presents on a descriptive and communicative
dimension the key elements that anyone should know about the monument (in an
understandable form, suitably subtitled for both adults and children).
4. Glossary: important words that are used at the text.
5. The 3D object (downloadable in a pdf file and in a simple online version).
6. Comprehension questions.
7. Activities: puzzles, hotspot images, games.
8. Quiz: a short quiz (usually 10–20 Questions) about the Monument. You need to score
70 % in order to pass.
9. References: resources and links for further studying
Fig. 3. Example of lessons’ structure. The 3D file is used for download purpose, encouraging
the schools to use it for 3D printing and then painting it in classrooms by the kids.
In every lesson, there is a time estimation, to help educators and users to manage
their own time and pace when studying about the monument. At the end, there is a
“Review” section. Every visitor can write a review and choose 1–5 stars for the course.
This way, courses with better score will be proposed at the Home Page. Finally, users
can print a “statement of accomplishment” of the Course, if they pass the Quiz.
The web site, as an educational tool, uses different digital heritage data depending
on the age of the users. For kids aged between 6 and 10 years old, the ortho-rectified
pictures were used as a puzzle game (Fig. 4). For older kids the same pictures can be
used as “find the hot-spot” game, were useful information is hidden under different hot-
spots of the image (Fig. 5). In all ages, 3D models of the Monument are used to encourage
interaction but - most important - to give the teacher the opportunity to download and
3D print the model. This way, the kids can touch, paint and interact with the Monument’s
model, enriching the overall learning experience. Furthermore, students could be divided
into groups (either e-course is a lesson inside the classroom or homework assignment)
Fig. 4. Using ortho-rectified pictures for developing interesting online games for educational
purposes. In this case, a picture was used for an online puzzle game, addressing kids aged 6 years
old
Fig. 5. Using photogrammetric or panorama pictures for developing interesting online games for
educational purposes. These pictures cover bigger areas than usual pictures and more “hot spots”
could be applied on them, with more information.
so to foster collaboration among them to produce shared ideas and form knowledge
through communication and interaction. What is also achievable via the e-learning plat‐
form is the collaboration between students from different schools, either Cypriot or
International. These collaborative learning experiences are available by the proper use
of such ICT tools.
At the same time, the content of every lesson is different, depending on the ages of
the users. For users aged over 16 years old, for life-learning users and for vocational
training, the content is more advanced; the glossary is more complex and the demands
(quiz, comprehensive questions) are higher. The courses (monuments) that are addressed
for younger children, the content is simplified according to their age. In order to organize
the courses and improve the user experience of the site, three virtual teachers were
developed. Each teacher has different ages in her/his virtual classroom and different
courses (monuments). This way, if an educator teaches i.e. elementary school, she/he
would follow the elementary school virtual teacher, using fast and efficient the correct
courses for the classroom.
5 Results - Conclusions
Today is obvious that we have a lot of digital data (especially from crowdsourcing
platforms like Europeana) with controversial information. If any student, educator or
simple user needs to learn or study the history of a monument, usually she/he ends up
(through google search) to huge number of information, text, picture (with no obvious
connection or order) and without the semantics or the metadata that they need to study
the Monument.
In order to achieve rigid results, we combined the Europeana and Open Source Data
on the Internet, with an educational tool. The model construction is based on the cross-
cultural approach, which responds to the multicultural features of present era but at the
same time to the contemporary pedagogical and methodological directions.
The result is an online learning user-friendly experience, where visitors (educators,
students and simple users) have the opportunity to read, watch, interact (3D model) and
test their knowledge in cultural heritage. The information is adjusted to their needs and,
if they finish the course successfully, they can use the references to deepen their knowl‐
edge.
The affordances described above: personalized content for different ages, roles
(teachers - students), professionals, can be furthered integrated with interdisciplinary
learning dimensions. Specific objects can be linked with external or internal data to
related sources from various relevant disciplines. An existed network could be broad‐
ened with historical, socio-cultural, economical and other geographical links, which
allow a stronger interpretation of the presented cultural artifact itself [10].
Moreover, one of the future features of the platform would be the ability to provide
a 3D model of the monument constructed by photographs taken by actual users being
on the site’s physical environment. For cultural e-courses this kind of personalized e-
learning functionality, and therefore personalized interaction, is one of the most impor‐
tant requirements [10].
Last but not least, Europeana’s content enrichment with Image Based Modeling Data
is considered to be vital in order to overcome the existing obstacles and limitations and
to provide the user with superior robusting experiences.
6 Future Work
An evaluation of the platform is planned to be conducted with (a subset of) our target
users participating at the procedure. Students, teachers as well as teaching inspectors in
two schools, one in Greece and the other one in Cyprus, will evaluate the platform. In
order to cover the cross-cultural aspect of the application, it is important to engage both
the users that know the monuments and can visit them on site (Cypriots) and the users
that might have some information about the monuments but they are not capable of
visiting those (Greeks). The evaluation of the educational aspect of the platform will
follow a methodology of joint techniques, in order to have a more integrated and overall
view of the evaluation results.
In this phase a prototype of the platform is completed and running successfully and
we intend to run a formative evaluation. Formative evaluation, is usually taking place
during the development or improvement of an educational program or product [11], in
order to make it more effective and make any changes appropriate as well as to check
whether it meets the learning needs of the users [12, 13]. It is a user-centered
approach [14].
Before evaluating the platform, some definitions of the terms involved must be clear.
First, educational software is considered any application used for the computational
support of education and teaching. Evaluation of educational software is the systematic
collection, analysis and interpretation of information of any aspect of a product, with
direction to the ascertainment of its effectiveness and efficiency or the estimation of any
other parameters relevant to its applicability [15]. Generally, in the evaluation process
of educational software, the educational environment where the software is used, the
perceptions of the students as well as other parameters involved in the learning process
(stakeholders, environment), should be taken into consideration [16].
Regarding the collection of research data, it has been noted that the persistence with
one technique can influence or misrepresent the image of the subject that the researcher
has dealt with [17]. So, it is necessary to use combinatorial methods, such as interview,
focus groups, and/or (e-)questionnaires. Those three methods are planned to be
employed in our evaluation process. The teaching inspectors will evaluate the platform
through interviews, teachers will use questionnaires and finally, students will evaluate
the software in focus groups. These pairs of method-user were assigned according to
the types of users and the quality/quantity of the data we want to receive from each one
type.
In detail, interview is particularly prevalent and effective in the case of ‘qualitative’
surveys. In this research, semi structured interviews on subjects that promote, teach and
communicate historical knowledge, are directed to ‘capture’ ideas, thoughts and repre‐
sentations that the subject has on the particular research topic [18–20]. The questions
will be categorized into: (a) demographic (general nature and on the expertise of
individuals), (b) cognitive (knowledge held by people on the subject), (c) experience
and (d) personal opinion [21]. Sampling and analytical issues of data resulting from the
interviews will be handled based on the international relevant literature [22].
The planned questionnaires will take account both specific ethical issues (informed
consent, understanding of the opportunities provided by the research, confidentiality,
anonymity and non-traceability, reliability and validity) and basic functionality criteria,
structure and usability. The aim is to create a short, focused and easy to use tool, which
will be distributed and (at the best case scenario) completed electronically through the
Internet [17].
Last but not least, students are the most important evaluators with the, most difficult
and intensive to design, evaluation technique. The “special group session” (focus group)
is a technique of qualitative data collection, in which a small number of participants
(usually 4–9 people) discuss and comment on a specific topic, under the guidance of a
trained mediator. Participants are distinguished by some common characteristics, such
as their relationship with the matter under investigation, but also by the growing dynamic
between them during the session [23]. The value of this technique is based on the inter‐
action and the ‘chemistry’ that can be developed between the participants, which
encourages them to express freely and with intimacy, unobstructed views, attitudes and
ideas they have about the subject matter, agreements or disagreements with the opinions
or attitudes of the other attendees [24]. All these are describing a full educational context
when learning is taking place inside the classroom. Such forms of rich interactions and
dialogs we are expecting to record from students.
Acknowledgements. 1. Europeana Space (http://www.europeana-space.eu/) has received

funding from the European Union’s ICT Policy Support Program as part of the Competitiveness
and Innovation Framework Program, under GA no. 621037.
2. ITN-DCH (http://www.itn-dch.eu/) project has received funding from the European
Union’s Seventh Framework program for research, technological development and demonstration
under GA no. 608013.
3. INCEPTION (http://www.inception-project.eu/) has received funding from the EU’s
H2020 Reflective framework programme for research and innovation under GA no. 665220.
4. 4DCH (http://www.4d-ch-world.eu/) project has received funding from the European
Union’s Seventh Framework program for research, technological development and
demonstration.
5. Lo-Cloud (http://www.locloud.eu/) is co-funded under the CIP ICT-PSP program.
6. ViMM (http://www.vi-mm.eu/) has received funding from the EU’s H2020 framework
programme for support and coordination actions under GA no. 727107.
References
1. Ott, M., Pozzi, F.: ICT and cultural heritage education: which added value? In: Lytras, M.D.,
Carroll, J.M., Damiani, E., Tennyson, Robert, D. (eds.) WSKS 2008. LNCS (LNAI), vol.
2. Stafford, T.: Teaching Visual Literacy in the Primary Classroom: Comic Books, Film,
Television and Picture Narratives. Routledge, London (2010)
3. Song, Y.: “Bring Your Own Device (BYOD)” for seamless science inquiry in a primary
school. Comput. Educ. 74, 50–60 (2014)
4. Afreen, R.: Bring Your Own Device (BYOD) in higher education: Opportunities and
challenges. Int J Emerg Trends Technol Comput Sci 3, 233–236 (2014)
5. Alberta Education: Bring your own device: a guide for schools (2012). Accessed 25 June 2012
6. Palmer research white paper: best practices for enabling BYOD in education (2013)
7. Bradford Networks: The impact of BYOD in education
8. Yuan, L., Powell, S.: MOOCs and disruptive innovation: implications for higher education.
eLearning Pap. In-depth. 33, 1–7 (2013)
9. Zutshi, S., O’Hare, S., Rodafinos, A.: Experiences in MOOCs: the perspective of students.
Am. J. Distance Educ. 27, 218–227 (2013)
10. Styliadis, A.D., Akbaylar, I.I., Papadopoulou, D.A., Hasanagas, N.D., Roussa, S.A., Sexidis,
L.A.: Metadata-based heritage sites modeling with e-learning functionality. J. Cult. Heritage
10, 296–312 (2009)
11. Postlethwaite, T.N.: Educational research: some basic concepts and terminology 1. Comp.
Educ. 1, 1–55 (2005)
12. Maslowski, R., Visscher, A.J.: Formative evaluation in educational computing research and
development. J. Res. Comput. Educ. 32, 239–255 (1999)
13. Reeves, T.C.: Evaluation of the design and development of IT tools in education. In: Voogt,
J., Knezek, G. (eds.) International Handbook of Information Technology in Primary and
Secondary Education, pp. 1037–1051. Springer, Heidelberg (2008)
14. Pintelas, P., Karatrantou, A., Panagiotakopoulos, C.: The evaluation of educational software
and its content. Deukaliοn (2012)
15. Panagiotakopoulos, C., Pierrakeas, C., Pintelas, P.: The Educational Software and Its
Evaluation. Metaichmio Publishing, Athens (2003)
16. Squires, D.: An heuristic approach to the evaluation of educational multimedia software
(1997). http://www.media.uwe.ac.uk/~masoud/cal-97/papers/squires.htm
17. Cohen, L., Manion, L., Morrison, K.: Methodology of Educational Research. Ekfrasi, Athens
(1997)
18. Patton, M.Q.: Qualitative Evaluation and Research Methods. SAGE Publications Inc.,
Newbury Park (1990)
19. Opie, C.: Research Procedures Doing Educational Research. A Guide to First Time
Researchers. Sage Publications, London (2004)
20. Rubin, H.J., Rubin, I.: Qualitative Interviewing: The Art of Hearing Data. Sage Publication,
Thousand Oaks (1995)
21. Wallen, N.E., Fraenkel, J.R.: Educational Research: A Guide to the Process. Psychology Press,
Cambridge (2001)
22. Mertler, C., Charles, C.: Introduction to Research. Pearson Education Inc., New York (2005)
23. Litosseliti, L.: Using Focus Groups in Research. A&C Black, London (2003)
24. Goldenkoff, R.: Using focus groups. In: Wholey, J., Hatry, H., Newcomer, K. (eds.) Handbook
of Practical Program Evaluation, pp. 340–362. JosseyBass, San Francisco (2004)
Time-Travelling with Mobile Augmented
Reality: A Case Study on the Piazza
dei Miracoli
Mihai Duguleana1(&), Raffaello Brodi2, Florin Girbacia1,

Cristian Postelnicu1, Octavian Machidon1, and Marcello Carrozzino2
1
Faculty of Mechanical Engineering, Department of Automotive and Transport
Engineering, University Transilvania of Brasov, 500036 Brasov, Romania
{mihai.duguleana,garbacea,cristian-cezar.postelnicu,
octavian.machidon}@unitbv.ro
2
PERCRO – TECIP, Scuola Superiore Sant’Anna of Pisa,
56017 San Giuliano Terme, Italy
{raffaello.brondi,m.carrozzino}@sssup.it
Abstract. This paper presents a new application in the field of cultural heritage,
allowing outdoor site exploration throughout different periods of time, based on
Mobile Augmented Reality (MAR) technology. The purpose of this research is
to allow a free interaction metaphor between users and heritage landmarks, and
to enrich their travel experience with important historic facts. We use
Metaio SDK to implement this concept within an Android application. We take
the specific case of the Leaning Tower of Pisa, the Cathedral and the Baptistery,
all key landmarks from Piazza dei Miracoli in Pisa, to prove the usefulness of
this paradigm. Five epochs are presented within the application, together with
key data about each of them. We assess the usability and engagement of this
application by conducting a study with 15 users. The results obtained from the
user evaluation show that the concept is not only valid, but also attracting to
most of the people. The findings suggest that this kind of applications may
attract more visitors while also enhancing their visiting experience.
Keywords: Cultural heritage Augmented Reality Mobile phones

UNESCO World Heritage User experience
1 Introduction
Today, Augmented Reality (AR) is widely used in several contexts including Cultural
Heritage. By combining real and virtual worlds, AR aims at enriching the user expe-
rience. The current research trends regard AR developments in the broader field of
tourism (including visiting historical landmarks and cultural heritage sites) as instru-
ments capable of providing extended information about the visualized items while also
being able to maximize the visitors’ satisfaction and enhance their experience [1].
There is a comprehensive research activity within the field of AR in museums, with
applications targeting the improvement of the visitor’s experience by creating an
informational synergy between the real and digital world, providing additional digital

DOI: 10.1007/978-3-319-48496-9_73
Time-Travelling with Mobile Augmented Reality 903
material (3D data, images, videos) about the cultural assets. The stated goal of many
heritage studies that use AR is to offer, by the means of digital restoration, a historical
accurate visual/audio perspective on landscapes, items and activities no longer avail-
able (either totally or partially).
There are many initiatives that use AR as a starting point, and build over this
technology. For example, in [2] is proposed the use of serious games (enhanced with
immersive technologies such as virtual environments and augmented reality) to
improve the cultural heritage experience, bringing it closer to the large public and
enabling learning cultural content in an engaging way. Other studies use the metaphor
of a “flash-light” to provide AR content [3], or 3D audio content as means to augment
the reality [4].
However, despite such a wide spreading of AR systems and guides in the field of
cultural heritage, time-based applications are scarcely exploited, and their influence on
the visitor’s behavior has not been properly evaluated so far. Ha, T. et al. proposed a
semantic model based on meta-data rather than actively using multiple planes for each
era [5]. In [6], authors use a display bar and 2D images, instead of 3D models, to
display the augmented content. Moreover, most of the literature focuses on indoor
applications (museum guides, virtual and augmented installations and exhibitions for a
museum). This paper proposes an AR-based time-travel guide and analyzes the
effectiveness of the system on visitors/tourists of the largest landmark from Pisa, the
Piazza dei Miracoli. Based on this analysis, we summarize a few guidelines for future
AR applications used as outdoor time-based guides, in order to raise their effectiveness
and improve the user interaction.
1.1 Related Works

There are a lot of studies and experimental implementations developed for integrating
virtual or augmented reality systems in cultural heritage sites.
In [1] the researchers present the implementation of a mobile-AR educational game
for iPhones destined to aid the visitor’s understanding about an archeological site and
exhibition. A flexible 3D semi-immersive platform for the virtual recreation of Cultural
Heritage monuments with educational applications is presented in [7]. The authors also
conducted a survey to evaluate the impact of the platform usage on students, and the
results showed that it had the potential to raise student interest in cultural heritage and
while also easing the learning process.
Immersive systems have a great potential in improving the learning process, and are
thus targeted as having possible applications in the cultural heritage field. Researchers
in [8] have created an information landscape-based immersive system intended to allow
a fast and easy navigation among a large volume of various types of cultural heritage
information (text, audio, and video).
Other works in this area [9–11] include the development of immersive tour
experience systems for cultural tour sites. These are basically observation posts
enhanced with audio/video augmented reality to provide the immersive experience by
superimposing virtual cultural heritage artifacts on their historical accurate position in
the real world (thus being able to virtually reconstruct lost landscapes, buildings, etc.).
904 M. Duguleana et al.
In [12] the visitor experience was also evaluated using a questionnaire, the users rating
it as positive and appreciating the novelty of the AR system.
Most recent research targets implementing augmented reality systems on mobile
devices, for raising their availability and ease-of-use [13]. A smartphone 3d model
visualization system based on augmented reality is presented in [14] as being a novel
approach for implementing such visualization systems for outdoor environments, such
as the one proposed in this paper.
Another example is the work described in [15], where a mobile application
designed to improve the user’s cultural experience during the sightseeing of a city of art
is described. This app basically detects the POIs inside the mobile device’s camera’s
video feed and composes an augmented scene with a virtual image slider, providing
thus the user with additional digital content regarding the specific item retrieved using a
cloud computing service.
There are several other related papers [16] that describe similar applications of
interactive augmented reality systems for enhanced exploration of cultural heritage
sites and artifacts. As mentioned above, all these papers present the design and
implementation of various AR/VR systems in cultural heritage environments but lack a
proper evaluation of how their usage impacts the visitors of such sites.
The authors of [17] are among the first to follow such a direction by showing the
potential role of embodied AR experience in changing thoughts and beliefs about a
certain place (from a cultural/social point of view). Also, [18] presents a MAR travel
guide and evaluates the relation between this system and the user’s emotion, adoption
behavior and other cognitive implications. Regarding the user adoption of such AR
systems, in [19] are identified three factors as being very influential in encouraging
cultural heritage visitors to use AR apps: technology readiness, visual factor of the AR
and situational factor.
2 Methods and Materials
2.1 MAR Setup

The AR application presented in this paper is called TowerAR. It is an Android based
application that uses MAR to present historical information (3D models and pho-
tographs) from Piazza dei Miracoli, Pisa. The application was developed using
Metaio SDK library (www.metaio.com). The MAR experience is provided via a
Samsung Galaxy Tab S2 tablet. In order to augment the video stream from the camera,
the user can point the tablet to a monument and can select the information related to it
that will be displayed registered on the target. Users can select to display just 2D
information on top of the video or the full 3D reconstruction of the building. The main
functionalities of the AR system are:
• AR view of 3D models: This is the main view of the TowerAR application where
the 3D models are displayed co-located with the real monument, based on the
selected time period. The view is presented in the Fig. 1.
Fig. 1. The TowerAR app main components
• Detailed information view: Shows detailed information as historical images dis-

played close to the selected monument. Each of the photographs contains a
description of the monument from the specific historic period.
• Timeline selection: Allows selecting different periods using specific buttons.
• Landmark selection: Allows selecting between the 3 different landmarks: the
Tower, the Cathedral or the Baptistery (Fig. 2).
Fig. 2. The 3D models of the Leaning Tower used for the time-based guide (years 1180, 1280
and 1360)
2.2 Application Architecture

TowerAR has a relatively simple design, which is flexible enough to be customized for
any heritage site with multiple landmarks (Fig. 3). The user can navigate through 3
separate panels. In the dashboard area, users can choose among the monuments which
are closer to their location. As soon as a monument is selected, the camera turns on and
the users can point the mobile device towards the monument which they want to learn
more about. Using a timeline bar, users can choose the time in history during which
they want to visualize the augmented stream. Based on the Metaio SDK, the mobile
device identifies the landmark and overwrite the output with information collected from
the 3D models and the content databases.
Fig. 3. TowerAR architecture
2.3 Case Study: Piazza dei Miracoli

The Piazza dei Miracoli is an UNESCO World Heritage Site located in center of Pisa,
Tuscany, Italy. This important landmark is regarded as one of the most beautiful
European medieval centers due to its fine architectural elements. The square consists of
4 religious structures: the leaning tower, the Cathedral, the Baptistry and the Sacred
Cemetery. The evolution of these edifices has been well documented over time. The
tower of Pisa, for example, was built in 3 stages, over a time span of 177 years [20].
Both the Cathedral and the Baptistery evolved a lot since they were first completed.
Given the extent of happenings that could influence the present form of the Piazza dei
Miracoli (starting with the Pisan wars, continuing with World War Two and Mus-
solini’s attempt to fix the tilt), it is truly a miracle the result is what we see today. This
gave us the idea of really showing to the tourists wandering the square the way the
tower changed over the centuries. The Cathedral and the Baptistery also have their own
stories. The Cathedral’s construction started in 1064, almost 120 years before the
construction of the tower begun. The building evolved over the next hundreds of years
to the form we can witness today. The Baptistery is the largest building of its kind in
Italy, and was not originally thought to sustain a dome. The most recent roof form was
given by Nicola and Giovanni Pisano, by the XIV century.
3 Evaluation
A user study has been conducted in order to assess the usability and appreciation of the
developed application. The system has been proposed to several tourists visiting Piazza
dei Miracoli, and the impressions of the participants have been collected using an
online questionnaire.
3.1 Procedure
The experimenters asked people visiting the square to participate in the experiment.
After an explanation of the purpose of the experiment and an introduction to the
TowerAR application, a tablet has been given to the participants who were left free to
visit the square. In order to get information about the surrounding monuments, they
only needed to point the tablet towards the points of interest. Participants were allowed
to use the tablet as much as they wanted (each participant used the app between 3 and
10 min). After they tried the application, participants were asked to fill a questionnaire
which collected their impressions.
3.2 Participants
The participants to the experiment have been recruited among tourists visiting the
square. A total of 15 subjects, 11 males and 4 females subjects, aged between 17 and 48
(31.14 ± 7.95) years took part at the experiment. During the recruitment we asked the
participants to rate their usage of mobile devices on a Likert-scale (from 1 to 7) and if
they have ever previously used an AR application. Each one of them uses a smartphone
or tablet almost every day (6.6 ± 0.71) but just a third of them have ever used or seen
an AR application.
3.3 Questionnaire Design

The questionnaire proposed to the users (Table 1) was composed by a first section
intended to collect some demographic information (gender and age) and the degree of
familiarity of the subjects with mobile devices and AR applications. The second section
(Q3–Q17) aims to assess the usability of the application using the Handheld Aug-
mented Reality Usability Scale (HARUS) [21]. The HARUS questionnaire has been
purposely developed to evaluate the comprehensibility and manipulability of Handheld
Augmented Reality (HAR) applications. Comprehensibility refers to the “ease of
understanding the information presented” by an HAR application. The manipulability
refers to “ease of handling the AR application as the user performs a task”. HARUS
questionnaire contains eight statements for the comprehensibility scale and eight
statements for the manipulability scale. We decided to remove one of the manipula-
bility questions (“I found it easy to input information through the application”) because
it is not relevant in the evaluation of the proposed application. The last section (Q18–
Q26) aims to assess the enjoyment and usefulness of the application. The questions
Table 1. Evaluation questionnaire

Personal Skills Q1: How often do you use mobile devices (tablet, smartphone)?
Q2: Have you experienced Augmented Reality before this test?
Comprehensibility Q3: I thought that interacting with this application requires a lot of
mental effort
Q4: I thought the amount of information displayed on screen was
appropriate
Q5: I thought that the information displayed on screen was difficult to
read
Q6: I felt that the information display was responding fast enough
Q7: I thought that the information displayed on screen was confusing
Q8: I thought the words and symbols on screen were easy to read
Q9: I felt that the display was flickering too much
Q10: I thought that the information displayed on screen was consistent
Manipulability Q11: I thought that interacting with this application requires a lot of
body muscle effort
Q12: I felt that using the application was comfortable for my arms and
hands
Q13: I found the device difficult to hold while operating the application
Q14: I felt that my arm or hand became tired after using the application
Q15: I thought the application is easy to control
Q16: I felt that I was losing grip and dropping the device at some point
Q17: I thought the operation of this application is simple and
uncomplicated
Enjoyment Q18: I enjoyed using the application
Q19: I found the application unpleasant
Q20: I found the application exciting
Q21: I found the application boring
Usefulness Q22: By using the app, I could quickly and easily find historical pictures
and information
Q23: By using the app, I learned more about the history of Piazza dei
Miracoli
Q24: By using the app, I could quickly find historical pictures and
information from places nearby
Q25: By using the app, I am more likely to find historical pictures and
information that interests me
Q26: I found useful being able to browse different historical ages with
the application
have been extracted from the questionnaires proposed by Haugstvedt and Krogstie [22]
using the “Perceived usefulness” and “Perceived enjoyment” constructs. Q26 has been
added in order to evaluate the feature which allows users to switch between different
augmented views linked to different epochs.
All the questions, except Q2 which is a True/False question, use a 7-point Likert
scale, ranging from 1 – “strongly disagree”, to 7 – “strongly agree”.
3.4 Results
The questionnaires results have been analyzed in order to evaluate the developed
application in terms of usability and appreciation. Following the instructions provided by
the authors in [21, 22], the answers have been aggregated in order to obtain a value for
each one of the four constructs defined by the two questionnaires: Comprehensibility,
Manipulability, Enjoyment and Usefulness. To compute the construct outcomes, for each
user, we first invert the results of negatively stated items and then we sum all the values
related to the specific construct. The obtained score is mapped to a range of 0 to 100.
Figure 4 shows the outcomes of the questionnaire for each of the questions. The
values reported here have not been inverted yet. The values obtained for each question
Fig. 4. The plot shows the outcome for each question proposed to the users. The mean and the
interquartile values are reported in the boxplots. Outliers are reported as asterisks
Fig. 5. The plot displays the results for the four constructs. The mean and the interquartile
values are reported in the boxplots. Outliers are reported as asterisks
Fig. 6. Using the TowerAR app
indicates a positive evaluation. This can be seen more clearly in Fig. 5 where the values
for each construct – Comprehensibility (M = 85.59; SD = 4.98), Manipulability
(M = 87.61; SD = 7.58), Enjoyment (M = 87.38; SD = 7.12) and Usefulness
(M = 79.42; SD = 7.89), are reported.
4 Conclusions and Future Developments
In this paper, we presented TowerAR, an application which allows MAR-based

time-travelling for the monuments of the Piazza dei Miracoli, Pisa. Historical infor-
mation (3D models and images) is visible when the users point the mobile device
towards the edifices. A preliminary user study was conducted to evaluate the com-
prehensibility, manipulability, enjoyment and usefulness of TowerAR. After plotting
the results, we have found a high score for all analyzed constructs. The results of
experiments show that users not only were satisfied by the application, but also felt
excited when using it (Fig. 6). Further developments can be made in improving the 3D
models, adding more locations, testing the application with a larger user base, imple-
menting an automatic selection based on GPS coordinates or linking the AR content to
web-based libraries.
Acknowledgements. This paper is supported by European Union’s Horizon 2020 research and
innovation programme under grant agreement No. 692103, project eHERITAGE (Expanding the
Research and Innovation Capacity in Cultural Heritage Virtual Reality Applications).
References
1. Angelopoulou, A., Economou, D., Bouki, V., Psarrou, A., Jin, L., Pritchard, C., Kolyda, F.:
Mobile augmented reality for cultural heritage. In: International Conference on Mobile
Wireless Middleware, Operating Systems, and Applications, pp. 15–22 (2011)
2. Mortara, M., Catalano, C.E., Bellotti, F., Fiucci, G., Houry-Panchetti, M., Petridis, P.:
Learning cultural heritage by serious games. J. Cult. Heritage 15(3), 318–325 (2014)
3. Ridel, B., Reuter, P., Laviole, J., Mellado, N., Couture, N., Granier, X.: The revealing
flashlight: interactive spatial augmented reality for detail exploration of cultural heritage
artifacts. J. Comput. Cult. Heritage 7(2), 1–18 (2014). Association for Computing
Machinery (ACM)
4. D’Auria, D., Di Mauro, D., Calandra, D.M., Cutugno, F.: A 3D audio augmented reality
system for a cultural heritage management and fruition. J. Digit. Inf. Manag. 13(4), 203–209
(2015)
5. Ha, T., Kim, Y., Kim, E., Kim, K., Lim, S., Hong, S., et al.: K-Culture time machine:
development of creation and provision technology for time-space-connected cultural
contents. In: Proceedings of the International Conference on Human Interface and the
Management of Information, pp. 428–435 (2015)
6. Fiore, A., Mainetti, L., Manco, L., Marra, P.: Augmented reality for allowing time
navigation in cultural tourism experiences: a case study. In: International Conference on
Augmented and Virtual Reality, pp. 296–301 (2014)
7. Bustillo, A., Alaguero, M., Miguel, I., Saiz, J.M., Iglesias, L.S.: A flexible platform for the
creation of 3D semi-immersive environments to teach cultural heritage. Digit. Appl.
Archaeol. Cult. Heritage 2(4), 248–259 (2015)
8. Ruffaldi, E., Evangelista, C., Neri, V., Carrozzino, M., Bergamasco, M.: Design of
information landscapes for cultural heritage content. In: Proceedings of the 3rd International
Conference on Digital Interactive Media in Entertainment and Arts, pp. 113–119 (2008)
9. Carrozzino, M., Bergamasco, M.: Beyond virtual museums: experiencing immersive virtual
reality in real museums. J. Cult. Heritage 11(4), 452–458 (2010)
10. Gaugne, R., Gouranton, V., Dumont, G., Chauffaut, A., Arnaldi, B.: Immersia, an open
immersive infrastructure: doing archaeology in virtual reality. Archeologia e Calcolatori,
Supplement 5, 1–10 (2014)
11. Malomo, L., Banterle, F., Pingi, P., Gabellone, F., Scopigno, R.: VirtualTour: a system for
exploring cultural heritage sites in an immersive way. In: 2015 Digital Heritage, vol. 1,
pp. 309–312 (2015)
12. Olsson, T., Kärkkäinen, T., Lagerstam, E., Ventä-Olkkonen, L.: User evaluation of mobile
augmented reality scenarios. J. Ambient Intell. Smart Environ. 4(1), 29–47 (2012)
13. Casella, G., Coelho, M.: Augmented heritage: situating augmented reality mobile apps in
cultural heritage communication. In: Proceedings of the 2013 International Conference on
Information Systems and Design of Communication, pp. 138–140 (2013)
14. Han, J.G., Park, K.W., Ban, K.J., Kim, E.K.: Cultural heritage sites visualization system
based on outdoor augmented reality. AASRI Procedia 4, 64–71 (2013)
15. Vecchio, P., Mele, F., De Paolis, L.T., Epicoco, I., Mancini, M., Aloisio, G.: Cloud
computing and augmented reality for cultural heritage. In: International Conference on
Augmented and Virtual Reality, pp. 51–60 (2015)
16. Seo, B.K., Kim, K., Park, J., Park, J.I.: A tracking framework for augmented reality tours on
cultural heritage sites. In: Proceedings of the 9th ACM SIGGRAPH Conference on
Virtual-Reality Continuum and its Applications in Industry, pp. 169–174 (2010)
17. Oleksy, T., Wnuk, A.: Augmented places: an impact of embodied historical experience on
attitudes towards places. Comput. Hum. Behav. 57, 11–16 (2016)
18. Kourouthanassis, P., Boletsis, C., Bardaki, C., Chasanidou, D.: Tourists’ responses to mobile
augmented reality travel guides: the role of emotions on adoption behavior. Pervasive
Mobile Comput. 18(7), 1–87 (2015)
19. Chung, N., Han, H., Joun, Y.: Tourists’ intention to visit a destination: the role of augmented
reality (AR) application for a heritage site. Comput. Hum. Behav. 50, 588–599 (2015)
20. Burland, J.B., Jamiolkowski, M., Viggiani, C.: The stabilization of the leaning tower of Pisa.
Japan. Geotechn. Soc. 43(5), 63–80 (2003)
21. Santos, M.E.C., Polvi, J., Taketomi, T., Yamamoto, G., Sandor, C., Kato, H.: Toward
standard usability questionnaires for handheld augmented reality. IEEE Comput. Graph.
Appl. 35(5), 66–75 (2015)
22. Haugstvedt, A.C., Krogstie, J.: Mobile augmented reality for cultural heritage: a technology
acceptance study. In: 2012 IEEE International Symposium on Mixed and Augmented
Reality (ISMAR), pp. 247–255 (2012)
Author Index
Aalil, Issam II-160 Brumana, Raffaella I-104

Adamec, Michael I-836 Brunetaud, Xavier I-681, II-53, II-160
Addabbo, Tommaso I-669 Bruno, Fabio I-269, I-805
Agapiou, Athos I-615, II-91, II-171, II-275 Bueno Benito, Gumersindo I-880
Agrafiotis, P. I-805 Bugadze, Nino II-16
Al-Mukhtar, Muzahim I-681, II-53, II-160
Amditis, Angelos I-50 Calamai, Silvia II-232
Anagnostoudis, Pavlos II-118 Calcerano, Filippo I-91
Andersson, John II-33 Caldeira, Altino Barbosa II-248
Andreoli, Roberto I-814 Caneva, Stefano G. II-222
Antoniou, Angeliki I-501 Caputo, Federico I-579
Arnaldi, Bruno I-184, I-432 Cardaci, Alessio I-241
Athanasiou, George I-50 Cario, Gianni I-269
Athanasiou, Vasilis I-442, I-891, II-256, Carrozzino, Marcello I-902
II-265 Casagrande, Pedro II-63
Atki, Mustapha II-160 Castrillo, David I-442
Attenni, Martina I-533 Castro, F. I-805
Avouris, Nikolaos M. II-199 Cawthorne, Douglas I-27
Cerra, Daniele II-275
Bachlitzanakis, Vasileios I-257 Cessari, Luciano I-91
Badreddine, Dalal II-160 Chaaba, Ali II-160
Balda, Felix I-281, I-397 Chatzigrigoriou, Pavlos I-144, I-442, I-891,
Balogh, T. I-805 II-256, II-265
Banfi, Fabrizio I-104, I-116 Cheng, Chin-Fang I-659
Baratin, Laura I-645, I-711 Cherkaoui, Khalid II-160
Barazzetti, Luigi I-104 Ciacci, Andrea II-130
Barbieri, Loris I-269 Cigna, Francesca II-91
Barreau, Jean-Baptiste I-184 Clément, Serain II-283
Bartolomei, Cristiana I-533 Constantinou, Vaso II-214
Bassier, Maarten I-79 Corolla, Angela I-814
Becattini, Federico I-781 Cozza, Alessandro I-269
Beck, Kévin I-681, II-53, II-160 Cozza, Marco I-269
Bertozzi, Sara I-645, I-711 Crosta, Giovanni II-91
Bessa, Maria I-758 Cuca, Branka I-615, II-91, II-275
Biliotti, Francesca II-232
Biscione, Marilisa II-99, II-275 Danese, M. II-99
Bokolas, Vassilis I-891, II-118 Davardoust, Sanaz I-727
Bonsma, Iveta I-590 Davis, Graham R. I-219
Bonsma, Peter I-590 Davis, Joseph G. I-491
Bourexis, Fotis I-442 de Andrade, Bruno Amaral I-825
Brandhorst, Susanne I-281, I-397 de Bouw, Michael II-141
Brcic, Ramon II-275 de Castrillo, Maria Costi I-442
Bremer, Thomas I-281, I-397 de Sena, Ítalo Sousa I-825
Brodi, Raffaello I-902 Dekeyser, Liesbeth II-141
914 Author Index
Del Bimbo, Alberto I-781 Hadjidemetriou, George I-79

Deleplancque, Remi II-160 Hadjimitsis, Diofantos G. I-615, II-91,
Demesticha, S. I-805 II-171, II-275
Dimitros, Konstantinos I-758 Hegel, Wolfgang I-477
Dinu, M. II-151 Henze, Frank I-477
Doulamis, Anastasios I-50, I-805 Hernández, José Luis I-590
Doulgerakis, Adam I-758
Drap, P. I-805 Iadanza, Ernesto I-590
Dubois, Samuel II-141 Ide-Ektessabi, Ari I-232
Duchemin, Wandrille I-184 Inzerillo, Laura I-869
Duguleana, Mihai I-902 Ioannides, Marinos I-442, I-891, II-256,
Dumont, Georges I-184 II-265
Dworak, Daniel I-292 Ioannidis, Charalabos I-50, I-63, I-442
Ioannou, Andri II-214
Eineder, Michael II-275 Ippolito, Alfonso I-533
El-Rifai, Ibrahim I-232 Irodotou, Romylos I-27
Enkleistriotis, Archimadrite Neophytos Ispas, A. II-151
I-442
Jabi, Wassim I-171
Faggiano, Armando I-814 Jacobs, Marc I-849
Farazis, Giorgos I-758 Janvier, Romain II-53, II-160
Fauzia, Luca R. I-241 Janvier-Badosa, Sarah I-681, II-53
Ferracani, Andrea I-781
Ferrari, Federico II-16 Kadar, Manuella I-749
Fidas, Christos A. II-199 Kai-Browne, Arie I-281, I-397
Fischer, H. II-151 Kalle-den Oudsten, Inge II-209
Fonseca, Bráulio II-63 Kanellos, Tassos I-758
Fort, Ada I-669 Karafylli, Christina I-758
Frattini, Paolo II-91 Karafylli, Maria I-758
Fresa, Antonella I-409 Katsaros, Thomas I-691
Freville, Tristan I-420 Kersten, Thomas P. I-521
Fusaro, Gabriele I-579 Kikillos, F. I-805
Kim, Youngmo I-131
Ganetsos, Theodore I-691 Koehl, Mathieu I-420
Gaugne, Ronan I-184, I-432 Kohlmeyer, Kay I-281, I-397
Georgiou, Eftichia I-758 Koryciński, Wojciech I-556
Georgopoulos, A. I-805 Kouhartsiouk, Demetris II-91
Gessner, Ursula II-275 Koukopoulos, Dimitrios I-3, I-792
Gigliarelli, Elena I-91 Koukopoulos, Zois I-3
Girbacia, Florin I-902 Koukoulis, Kostas I-792
Gizzi, F.T. II-99 Kountouriotis, Vassilios I. I-758
Gómez-García-Bermejo, Jaime II-25 Koutsabasis, Panayiotis I-350
Gonnella, Julia I-397 Kovačević, Sandra II-72
Gordea, Sergiu I-309, II-109 Krauss, Thomas II-275
Goren, Arian I-281 Kriz, Pavel I-836
Gouranton, Valérie I-184, I-432 Kröber, Cindy I-362, I-477
Grellert, Marc I-39 Kubišta, Jiří II-191
Grussenmeyer, Pierre I-420 Kuprashvili, Nana II-16
Guasch-Jané, Maria Rosa I-737 Kuroczyński, Piotr I-292
Author Index 915
Kyriakidis, P. I-805, II-91 Marttila, Sanna I-309

Kyriazanos, Dimitris M. I-758 Masini, Nicola II-99, II-275
Medici, Marco I-590, II-16
La Guardia, M. II-41 Meliva, Tamar II-16
Lagudi, Antonio I-269 Merodo, José Antonio Fernandez II-91
Lai, Yu-Kun I-219 Mills, David I-219
Lampropoulos, Vassilis I-758 Moretti, Elvio I-645, I-711
Landucci, Lea I-781 Motos, Dionisis I-758
Lanitis, Andreas II-240 Motta, Silvio Romero Fonseca II-248
Larsson, Carolina II-10 Moura, Ana Clara Mourão I-825, II-63
Lasaponara, Rosa II-275 Mugnaini, Marco I-669
Lee, Jaeyong I-131 Münster, Sander I-362, I-477
Lee, Tung-Ming II-81 Murtiyoso, Arnadi I-420
Lehmann, Dennis I-397 Muzzupappa, Maurizio I-269
Lepouras, George I-501
Lerones, Pedro Martín I-590, II-25 Naudet, Yannick I-501
Leventis, Georgios I-442, I-891, II-256, Nicolas, Théophane I-432
II-265 Nigdelis, Pantelis II-118
Liarokapis, F. I-805 Nikolakopoulou, Vasiliki I-442, I-891,
Lindgren, Stefan II-10 II-256, II-265
Lindstaedt, Maren I-521 Nikolov, Ivan I-15
Liu, Chang I-219 Nisantzi, Argyro II-275
Llamas, Jose II-25 Nolè, Gabriele II-275
Lloyd, James I-859 Norton, Charles I-219
Lo Turco, Massimiliano I-579, I-600
Lobovikov-Katz, Anna I-337, I-880 Omelanowsky, Daniel I-521
Loizides, Fernando II-214 Osello, Anna I-727
Loog, Marco I-322 Ovidiu, Domsa I-749
Lukačević, Igor I-691
Lupia, Marco I-269 Pachta, Vasiliki I-199, I-701
Lykourentzou, Ioanna I-501 Panayiotou, Stavros II-240
Lysandrou, Vasiliki II-91, II-275 Panzardi, Enza I-669
Papaefthymiou, Margarita I-257
Machidon, Octavian I-902 Papageorgiou, Eirini I-442, I-625, I-891,
Madsen, Claus I-15 II-256, II-265
Mahgoub, Hend I-232 Papagiannakis, George I-257
Mahjoubi, Rachida II-160 Papagianni, Alexandra I-758
Maiellaro, Nicola II-178 Papayianni, Ioanna I-199, I-701
Maietti, Federica I-590 Parenti, Silvia I-590
Malandrino, Delfina I-814 Paterakis, Manolis I-758
Maly, Filip I-836 Pecci, A. II-99
Mancini, Matteo Flavio I-377 Pelcer–Vujačić, Olga II-72
Marcos Sánchez, Vanesa I-880 Peluso, Raffaele I-269
Margonis, Christos I-758 Petriaggi, B. Davidde I-805
Margottini, Claudio II-91 Pezzatini, Daniele I-781
Markiewicz, J. II-3 Pfarr-Harfst, Mieke I-39, I-477, I-544
Maroglou, Christos I-758 Pirozzi, Donato I-814
Martins, Joao I-880 Plesch, Sebastian I-281, I-397
916 Author Index
Popescu, C. II-151 Talma, Jean-Daniel I-184

Postelnicu, Cristian I-902 Tamborrino, Rosa I-727
Potamianos, Iakovos I-171 Tang, Jiexiong I-322
Potenza, M.R. II-99 Tavernier, Cédric I-432
Poullis, C. I-805 Thanos, Giorgos Konstandinos I-758
Pozzebon, Alessandro I-669, II-130, II-232 Themistocleous, Kyriacos I-513, II-91,
Pramartha, Cokorda I-491 II-171, II-275
Prechtel, Nikolas I-362, I-477 Theodorou, Ino-Eleni I-758
Thomopoulos, Christina Phobe I-758
Quattrone, Giuliana I-157 Thomopoulos, Stelios C.A. I-758
Truyen, Frederik I-409
Rădulescu, H.C. II-151 Tsimpiridis, Panagiotis I-758
Radvan, R. II-151 Turillazzi, Beatrice I-590
Ranaldi, Mirta I-814 Tzouvaras, Marios II-91
Raptis, George E. II-199
Ritacco, Gerardo I-269 Ubik, Sven II-191
Rizvic, S. I-805 Uhlemann, Rainer I-477
Rocha, Nicole II-63 Uricchio, Tiberio I-781
Roedel, P. II-151
van Biessum, Harry II-109
Rosin, Paul L. I-219
Van Damme, Peter II-141
Rossato, Luca I-567
Van der Have, Joeri-Aleksander II-141
Roșu, G. II-151
van der Lubbe, Jan C.A. I-322
Rozanska, Anna I-556
van Oppen, Branko F. II-222
Van Roy, Nathalie I-79
Saccuman, Roberto I-711 Vanderschoot, Joke I-849
Safa, Etienne I-184 Vanhellemont, Yves II-141
Sakali, Ayda II-160 Varasano, Antonietta II-178
Salvatore, Marta I-377 Vassilakis, Costas I-501
Santagati, Cettina I-600, I-869 Vergauwen, Maarten I-79
Santangelo, Gianluca I-814 Vermeir, Simon II-141
Scaduto, M.L. II-41 Versaci, Antonella I-241
Scarano, Vittorio I-814 Verstrynge, Els I-79
Scavone, M. II-99 Verykokou, Styliani I-50, I-63, I-442
Schindler, Alexander II-109 Vignoli, Michela I-309
Schreier, Gunter II-275 Vignoli, Valerio I-669
Schroyen, Morien I-849 Vosinakis, Spyros I-350
Scianna, A. II-41 Vouronikou, Nikoleta II-118
Sena, Ítalo II-63
Sileo, M. II-99 Wallace, Manolis I-501
Skarlatos, D. I-805 Wang, Tsung-Juang I-387
Skroumpelou, Katerina I-758 Wefers, Stefanie I-544
Soile, Sofia I-63, I-442 Weller, H. I-362
Sojref, Dalik I-880 Wu, Tsung Chiang I-209
Sovis, Christian I-442, I-891, II-256, II-265
Spizzichino, Daniele II-91 Yang, Wun-Bin I-387
Strand, Eva Andersson II-10 Yen, Alex Ya-Ning I-659, II-81
Strippgen, David I-281 Yen, Ya-Ning I-387
Author Index 917
Zacharakis, Dimitris I-758 Zeng, Yuan I-322

Zalama, Eduardo II-25 Zikas, Paul I-257
Zalonis, Andreas I-758 Ziri, Anna Elisabetta I-590
Zawieska, D. II-3

(Lecture Notes in Computer Science 10058) Marinos Ioannides, Eleanor Fink, Antonia Moropoulou, Monika Hagedorn-Saupe, Antonella Fresa, Gunnar Liestøl, Vlatka Rajcic, Pierre Grussenmeyer (eds.)-Digital.pdf

Uploaded by

Copyright:

Available Formats

(Lecture Notes in Computer Science 10058) Marinos Ioannides, Eleanor Fink, Antonia Moropoulou, Monika Hagedorn-Saupe, Antonella Fresa, Gunnar Liestøl, Vlatka Rajcic, Pierre Grussenmeyer (eds.)-Digital.pdf

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Lecture Notes in Computer Science 10058) Marinos Ioannides, Eleanor Fink, Antonia Moropoulou, Monika Hagedorn-Saupe, Antonella Fresa, Gunnar Liestøl, Vlatka Rajcic, Pierre Grussenmeyer (eds.)-Digital.pdf

Uploaded by

Copyright:

Available Formats

What is the purpose of this document?

What is the purpose of this document?

What topics are covered in the document?

What topics are covered in the document?

Marinos Ioannides · Eleanor Fink

Antonia Moropoulou · Monika Hagedorn-Saupe

Progress in Cultural Heritage:

Antonia Moropoulou Monika Hagedorn-Saupe

Antonella Fresa Gunnar Liestøl

Vlatka Rajcic Pierre Grussenmeyer (Eds.)

ISSN 0302-9743 ISSN 1611-3349 (electronic)

Library of Congress Control Number: 2016956494

© Springer International Publishing AG 2016

Printed on acid-free paper

This Springer imprint is published by Springer Nature

September 2016 Marinos Ioannides

Paper Review Chair

Local Organizing Committee

Agapiou, Athos Nobilakis, Elias

International Scientiﬁc Committee

De Niet, Marco Netherlands Liestøl, Gunnar Norway

Tsapatsoulis, Nicolas Cyprus Vermeulen, Frank Belgium

Dr. Ioannis A. Eliades,

Full Paper: 3D Reconstruction and 3D Modelling

Implementation and Usage Scenarios of a Participatory Platform

Benchmarking Close-range Structure from Motion 3D Reconstruction

Proportional Systems in the Design of the Cathedral of St. George

The Reconstruction – Argumentation Method: Proposal for a Minimum

Multi-scale 3D Modelling of Damaged Cultural Sites: Use Cases

Low Cost Technique for Accurate Geometric Documentation of Complex

Full Paper: Heritage Building Information Models (HBIM)

Implementation of Scan-to-BIM and FEM for the Documentation

Implementation Analysis and Design for Energy Efficient Intervention

Historic BIM in the Cloud . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

Building Information Modelling – A Novel Parametric Modeling Approach

Full Paper: Innovative Methods on Risk Assesment, Monitoring

Applicability of 3DVE Indicator for Cultural Heritage Landscape

Taking the Next Step in Digital Documentation of Historic Cities:

Risk Analysis and Vulnerability Assessment of Archeological Areas

Full Paper: Intangible Cultural Heritage Documentation

Parameterizing the Geometry and Visualizing the Lighting Method

Digital and Handcrafting Processes Applied to Sound-Studies

Full Paper: Digital Applications for Materials’ Preservation

Evolution of Building Materials and Philosophy in Construction: A Process

A Study of 3D Digital Simulation Analysis of Fire Charring Degree

Full Paper: Non-destructive Techniques in Cultural Heritage Conservation

Recovering Historical Film Footage by Processing

Multi-spectral Imaging System (IWN) for the Digitization and Investigation

Diagnostic Activities for the Planned and Preventive Conservation

Full Paper: Visualisation, VR and AR Methods and Applications

A Mobile, AR Inside-Out Positional Tracking Algorithm, (MARIOPOT),

Virtual and Augmented Reality Tools to Improve the Exploitation

Interacting with Simulated Archaeological Assets . . . . . . . . . . . . . . . . . . . . 281

Virtual Reconstruction 3.0: New Approach of Web-based Visualisation

CultureCam: An Interactive Search Tool for Small Image Galleries . . . . . . . . 309

Learning Algorithms for Digital Reconstruction of Van Gogh’s Drawings . . . 322

Full Paper: Digital Cultural Heritage in Education, Learning and Training

“Human” Technology in the Digital Era: Freehand Images and Analysis

Researching Knowledge Concerns in Virtual Historical Architecture . . . . . . . 362

Project Paper: Data Acquisition, Process and Management