TAMIL INSCRIPTION IN

PYTHON
 ABSTARCT
 Recognition of ancient Tamil characters has always been a challenge for Epigraphers.

 This is primarily because the language has evolved over several centuries and the character set
has over this time both expanded and diversified.

 This proposed work focuses on improving optical character recognition techniques for ancient
Tamil script which was in use between the 7th and 12th centuries.

 While comprehensively curating a functional data set for ancient Tamil characters is an
arduous task, in this work, a data set has been curated using cropped images of characters
found on certain temple inscriptions, specific to this time period as a case study.

 After using Otsu thresholding method for binarization of the image, a two-dimensional
convolution neural network is defined and used to train, classify, and recognize the ancient
Tamil characters.
 INTRODUCTION

 Inscriptions and manuscripts are important sources of information to understand the history
and culture of ancient civilizations.

 Inscriptions are ubiquitous in ancient Indian temples.

 They can be found on everything - from rocks, slabs, and pillars, to the walls of the temples.

 Most of these inscriptions convey vast and useful monarchical information about proceedings
of administrative and religious processes.

 These inscriptions are valuable documented proof to understand better the quality of life
during the specified era.

 Tamil Nadu tops the list in the Survey list of Survey of Indian Epigraphy (1996). This implies
that Tamil Nadu has a significantly large number of inscriptions. Tamil is an ancient language
of the world, and amongst the earliest languages in the Indian subcontinent.
 EXISTING SYSTEM

 Many organizations and institutions are involved in digitizing and documenting Tamil
inscriptions.

 This often includes creating digital copies of inscriptions, recording metadata, and organizing
databases for easy access.

 Some projects focus on creating databases specifically for Tamil inscriptions.

 These databases serve as repositories of information related to the inscriptions, including

their location, content, historical context, and translations.

 High-resolution photography and imaging techniques are commonly employed to capture

detailed images of Tamil inscriptions.

 This helps in preserving the inscriptions and facilitates further study and analysis.
 DISADVANTAGES

 Existing systems may be built on outdated technology, making it challenging to keep up with
modern standards and security requirements.

 Legacy systems may face difficulties integrating with newer technologies, hindering
interoperability and efficiency.

 Older systems may lack proper documentation, and support may be limited or unavailable,
making it challenging to troubleshoot issues.
 PROPOSED SYSTEM

 Develop a digital archiving system to store high-resolution images and detailed

documentation of Tamil inscriptions.

 Include metadata such as location, date of discovery, historical context, and associated
artifacts.

 Integrate linguistic analysis tools to study the language used in inscriptions and identify
linguistic patterns and changes over time.

 Support collaboration with linguists and epigraphists for in-depth language research.
 ADVANTAGES

 The system facilitates the digital preservation of Tamil inscriptions, ensuring that high-
resolution images and detailed documentation are securely stored for future generations.

 Regular updates keep the system aligned with the latest technologies, security standards, and
improvements in linguistic analysis tools, ensuring its long-term relevance and effectiveness.
SYSTEM REQUIREMENTS
HARDWARE REQUIREMENT

SOFTWARE REQUIREMENT
 HARDWARE REQUIREMENT

 System : Pentium i3 Processor

 Hard Disk : 500 GB

 Monitor : 15’’ LED

 Input Devices : Keyboard, Mouse

 Ram : 4 GB
 SOFTWARE REQUIREMENT

 Operating system : Windows 10.

 Coding Language : Python 3.8+

 IDE : Pycharm
ARCHITECTURE
 ALGORITHM

H5 MODEL ALGORITHM

 H5 is a file format to store structured data, it's not a model by itself.

 Keras saves models in this format as it can easily store the weights and model configuration in
a single file.

 The H5 Model addresses the mental and physical health issues related to trauma suffered by
refugees, the relationship between mental and physical health problems prevalent among
refugee populations, the potential for trauma to persist in refugee camps, and the need for a
new, more comprehensive model of refugee care.

 H5 files are organized into groups, which are similar to folders.

 Groups can contain datasets or other groups info, forming a hierarchical structure.
 MODULE DESCRIPTION

 INSCRIPTION DATASET CREATION MODULE

 MODEL TRAINING MODULE

 IMAGE PROCESSING MODULE

 INSCRIPTION PREDICTION MODULE

INSCRIPTION DATASET CREATION MODULE

 "Inscription dataset creation module" in the field of machine learning or data science.

 If this term is specific to a particular software module or tool developed after my last update,
I recommend checking the latest documentation, resources, or online materials associated with
that specific tool for an accurate and up-to-date definition.

MODEL TRAINING MODULE

 A "Model Training Module" typically refers to a component or module within a machine

learning framework or software that is responsible for the training phase of a machine
learning model.

 In the context of machine learning, training is the process of teaching a model to make
accurate predictions by learning patterns and relationships from a labeled dataset.
IMAGE PROCESSING MODULE

 An "Image Processing Module" refers to a component or module within a software system,

library, or framework that is designed to perform various operations and manipulations on
digital images.

 Image processing modules are commonly used in computer vision, computer graphics, and
other fields to enhance, analyze, or extract information from images.

INSCRIPTION PREDICTION MODULE

 An "Inscription Prediction Module" typically refers to a component or module within a

software system, often based on machine learning or artificial intelligence, that is designed to
predict or analyze inscriptions.

 This type of module may be part of a broader system dedicated to cultural heritage
preservation, historical research, or archaeological analysis.
 CONCLUSION

 By using a CNN and Image Recognition techniques, an operable system was designed for
modern and ancient Tamil.

 The difference in the style of the ancient Tamil scripts from the modern Tamil script posed as
a challenge to execute the task efficiently.

 Multiple samples of ancient inscriptions from some historical temples were taken as case
studies to implement the developed methodology.

 An acceptable accuracy rate of 77.7% was attained for these samples.

 The outputs obtained could not be digitally segmented due to lack of availability of any
language parser for ancient Tamil scripts. OCR techniques for ancient Tamilscripts is a rich
research topic.
code