INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE

Arslan et al., Vol.8, No.4, pp.88-113

Fingerprint Forensics in Crime Scene: A

Computer Science Approach

Bilgehan Arslan, Seref Sagiroglu

Department of of Computer Engineering, Gazi University, Ankara, Turkey

E-mail: {bilgehanarslan,ss}@gazi.edu.tr

ORCID ID: 0000-0002-5160-4408, 0000-0003-0805-5818

Research Paper Received: 28.11.2019 Revised: 23.12.2019 Accepted: 25.12.2019

Abstract—Biometry is essential when it comes to personal identification and verification in forensic applications, and the fingerprint
is one of the most commonly used biometric features in the field of biometric identification. To detect fingerprints at the crime scene,
to collect and transfer fingerprints in a way not to modify its nature, to make appropriate enhancements without causing loss of data
on fingerprints, to preserve and store fingerprints are complex processes. Appropriate realization of these processes is essential
for crime scene investigation. This study analyses and evaluates the crime scene investigation in a detailed way from a computer
science perspective with a special focus on fingerprints, and whole process is divided into three components. In light of the studies
conducted in the literature, 14 main problems that are encountered in the realization of these three components have been identified
and criticized. Afterwards, the disciplines supported by forensic science, and the components that compose the crime scene are
evaluated on the basis of four basic challenges encountered in the processes. Finally, to overcome 14 problems and four challenges,
six technology-supported solutions are presented.

Keywords—Fingerprint, crime scene investigation, forensic science, evidence, biometrics, computer science approach.

1. Introduction affecting people of the incident and people affected

by the incident. Forensic data is used for identifying
The crime scene is defined as the area where a the nature and reason of the incident, the people
crime has been committed, continued and spread. who affect the incident, people who are affected by
In the time interval between the beginning and end the incident, and external factors. These data are
of the crime, all the environments in which the collected as a result of CSI.
suspect and the victim are found are considered
crime scenes. For this reason, it contains a lot of CSI is based on scientific reasoning. This form of
forensic data about the crime committed. A crime research means that the CSI is systematic, method-
scene can be depicted as any physical environment ological, and logical. The first response to the crime
that could provide potential evidence to an investi- scene, securing the crime scene, documenting the
gator [1], [2]. The most general definition of crime crime scene, packing and preserving the materials
scene investigation (CSI) can be expressed as the obtained from the crime scene, collecting physical
analysis of the relationship between crime scene, evidence, developing hypotheses about the actions

88
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

taken at the crime scene, and reconstruction of crime be determined in advance. Thus, it is necessary to
are parts of the systematic reasoning process [1], have the capacity to evaluate all kinds of crime
[3]. scenes. Another component is Forensic Science.
Forensic sciences need to be very successful in
Forensic sciences have been guiding critical deci-
solving crimes, saving innocent people from the
sions made by governments, law enforcement of-
investigation process, and bringing criminals to jus-
ficers, scientists, lawyers, judges and technicians,
tice. Some practices in CSI require the collaboration
insurers, regulators, journalists, and even the public.
of different areas of expertise.
As well as specialists working in different dis-
ciplines of forensics, different segments of soci- For this reason, forensic science has a multidis-
ety have different perspectives and interpretations ciplinary structure. There is a growing need for
about what forensic science can do or cannot do. mechanisms and forensic science strategies that pro-
These perspectives affect the justice systems of vide accurate and reliable evidence within minutes.
societies. The ethical status of forensic science has Another important component of CSI is the human
changed dramatically in recent years. In the 1970s factor. Personnel working both at the crime scene
and 1980s, new scientific methods and techniques, and in the laboratory have a direct influence on the
devices, tools, technologies, standards, procedures, CSI process. No matter how major the development
and methodologies are used to improve the quality in forensic sciences is, the expert factor continues to
of assessments made today to overcome the inaccu- play an important role in decision-making in most
racies and deficiencies in evidence analysis. forensic processes. Forensic work often involves
interpretation and subjectivity. For this reason, cog-
Over the past twenty-five years, forensic sciences
nitive and human factors are important for making
have made a big leap thanks to scientific break-
accurate forensic decisions.
throughs and the rapid development of technology.
However, most of the studies focusing on CSIs The purpose of this paper is multifold:
have been done about procedures to be applied in
• To ensure reliable and detailed investigation of
the management of the process, staff management,
the crime scene,
evaluation of biological and biometric evidence, and
• To maximize the quality of Forensic data col-
assessing the compliance of the whole process with
lection and recovery procedures,
legal criteria. There are many comprehensive studies
• To propose a systematic approach to ensure that
focusing on identification, collection, and evaluation
the collected Forensic data is validated,
of fingerprint, and the legal validity of these steps.
• To encourage a more consistent methodology,
However, there are not many studies in which the
thus enabling more verifiable results to be pro-
development of technological infrastructures used
duced,
for investigation processes is evaluated from a com-
• To facilitate the transfer of valid evidence be-
puter science perspective.
tween crime scene investigator and law enforce-
One of the main components of the CSI process ment,
is the crime scene. The realization process and • To establish a framework of standards, princi-

components of the crime, the place where the crime ples, and approaches for the detection, record-
took place, the conditions in which the crime was ing, and recovery of forensic evidence at the
committed and those involved in the crime cannot scene.

89
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

This study focuses on the fingerprint. The CSI appropriately have a direct impact on the success
process is divided into three components: infor- of the criminal analysis process. The purpose of
mation acquisition, field expert activity, and tech- the CSI is to analyze and document the crime
nical evaluation. The relationship between each scene, to collect physical evidence, and ultimately
component and their relationship with fingerprint to reconstruct the incident [5]. To carry out the
are evaluated. In these components, the difficulties steps above, teamwork must be ensured. For this
encountered in the process steps applied to the reason, people who have different expertise should
fingerprint and the technical/technological problems work in sequence or together during these steps. All
are examined from a computer science perspective the forensic data that are found to be suitable for
and solutions are presented. In the second section of evaluation at the scene are transferred for a holistic
the study, CSI components are explained in detail. analysis to produce a definite result [5].
In the third section, the fingerprint collected from The CSI process is suitable for classification and
the crime scene is evaluated in terms of crime scene analysis with different approaches since it incor-
components. In the fourth section, the problems porates many components from various disciplines
encountered in the crime scene components in terms under one roof. Many studies have been carried
of fingerprints are determined, and in the fifth out so far; crime scenes have been evaluated from
section, the approaches that need to be developed the location of the scene, the type of evidence,
as solutions to these problems are presented. In the the size of the crime scene and the type of crime.
last section, the results of the study are given. Special disciplines such as forensics, psychology,
sociology, technology, criminology, biology have
2. Crime Scene Components been used for these evaluations [6]. In this study, the
crime scene process is considered from a computer
CSI requires science, logic, and law to make
science perspective, and the whole process was di-
joint decisions under one roof. The scene should
vided into information acquisition component, field
be evaluated in detail with all conditions, and all
expert activity component, and technical evaluation
kinds of physical evidence should be associated with
component. These components were analyzed using
the statements of the suspect and affected people,
critical key elements shown in Figure 1 [4], [6]–[8].
which turns CSI into a long and exhausting process
[4]. Also, since each crime scene has its specific To enlighten a forensic case successfully, critical
features, the steps taken in assessing a crime scene key elements need to be provided. These basic
and collecting evidence cannot be directly adapted elements are as follows [7]–[14]:
to different crime scenes. • Validity: The validity of the outcome of the

The investigation process usually starts at the CSI depends on the information collected by the
crime scene. The crime scene is defined as the research specialist about the case, the methods
place where the incident takes place directly, the and techniques used in the investigation of the
immediate environment of the crime scene and the scene, and the results of the analysis in the
remote environment that indirectly affects the crime laboratory. If these processes are considered as a
scene [5]. Actions of the first arriving team, such chain of interconnections, one’s small careless-
as the accurate identification of the crime scene ness or error can lead to misinterpretation of the
and protection, isolation and securing the evidence case, the misguidance of legal authorities, loss

90
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

time spent on an investigation is seen as an

extra cost for law enforcement officers in terms
of the employing expert and shift. However, an
inattentive analysis and a wrong result have a
devastating effect on the whole process. For this
reason, the research specialist should employ a
careful and thoughtful approach and produce
several different theories of crime based on
the information available after the necessary
screening. The time taken to produce reasonable
conclusions about the action is crucial.
• Cost: Forensic analytical procedures applied
Fig. 1: Key Elements of Crime Scene Component to the types of physical evidence can be very
costly. In a crime scene, evidence that is unre-
lated to the incident, or the analysis of evidence
of evidence, or even closure of the case due to that becomes unrecognizable even if it is related
lack of evidence. to the incident reduces the efficiency of the
• Integrity: The boundaries of the crime scene investigation and increases its cost. Also, at
can be of different dimensions depending on the some crime scene, specific experts may need
type and progress of the incident. The crime to work with specific equipment. Such require-
scene shows the way the incident is handled ments increase the cost of the investigation
and the relationship between the victim and process.
the perpetrator dynamically. The crime scene All process steps implemented in the crime scene
contains many traces and evidence about the components must be compatible with four key el-
perpetrator and victim, as well as the evidence ements. Also, specific criteria have been offered
of people who are not involved. It can be very in this study to evaluate each component within
difficult to associate the collected data with the itself and to define the relationship between some
crime, especially if the environment is not well components. In the third and fourth sections of
isolated during the first intervention to the scene the study, a detailed explanation is given on these
or if the crime occurs in public areas where criteria.
many people have access. The most important
point here is to detect the data related to the 2.1 Information Acquisition
crime directly and make accurate inferences
from the chain of custody. CSI is a comprehensive process that requires
• Time: CSI is a difficult and time-consuming the simultaneous or sequential implementation of
work, and in many ways, all staff working on complex assessment mechanisms, the use of a wide
the scene are running against time. The content range of equipment during the assessment steps,
of the evidence may deteriorate in time or and the expert’s work. The main purpose of this
may disappear, and the criminal may escape or process is to collect information to analyze the
involve in further actions of crime. Also, the nature of the crime, the sphere of influence, and

91
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

Fig. 2: Evidence collected during the CSI process [15], [16]

form a hypothesis that can revitalize the incident. process, different experts and disciplines are used
For this reason, all the steps in the CSI, from the to determine the procedures to be applied for the
reconstruction of the incident to the outcome of the collected evidence and to evaluate the quality of the
court, must be directly related to the information evidence [17].
collected from the scene.
All the statements, comments, documents, and 2.2 Field Expert Activity
marks that come from different sources and any-
thing containing information are called evidence. CSI has been considered as the most important
Any material or physical object that can be touched, aspect of a criminal investigation by many people.
alive or inanimate is called physical evidence. After High-quality and valid evidence for accurate and
being processed in laboratories, it can be used as fair results can only be obtained if the incident is
evidence in court for the reconstruction of the crime. examined professionally and effectively. During the
The statements of witnesses and accomplices are investigation of the crime, many different units can
called oral evidence. In addition to this, digital or work simultaneously.
concrete documents gathered from the scene are A team leader is assigned to each scene in order
expressed as documentary evidence. This evidence to direct and manage the investigation, to distribute
serves useful in making sense of the incident. Spe- the tasks to the units and ensure the correlation of
cialized staff and disciplines have to ensure the the units working with each other to investigate the
effective collection, preservation, packaging, and incident. The team leader cannot leave the scene
transport of evidence. After the data collection from the beginning to the end of the incident. The

92
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

leader’s main task is to manage and coordinate the evidence such as blood is collected directly and the
entire process [18]. non-visible ones such as latent fingerprint are tried
The staff who are assigned to detect the incident to be made visible by expert intervention and techni-
is defined as the initial responding officers. They cal infrastructure support. Finally, once the evidence
arrive at the scene first and ensure the safety of the has been made visible, if necessary, it is moved from
environment. This team consists of people working the scene to the laboratory for examination. For the
in law enforcement officers and they are specialized collected evidence, a more detailed examination is
in the type of incident such as narcotics, counter- carried out in the laboratory. These examinations are
terrorism, etc. Their mission is to maintain the mostly chemical and computer-aided applications to
original state of the scene, protect the evidence, and understand the content of the evidence or to make
to provide information flow to the next teams [18]. sense from the evidence. Due to scenarios such as
re-examination of evidence by the court or objection
Depending on the location and type of incident,
to a court decision, all evidence must be stored.
some people who have proven expertise such as
For this reason, many technical infrastructures are
anthropologists, blood pattern analysts, bomb tech-
needed to collect, transfer, evaluate, match and even
nicians, criminalists, fingerprint analysts, engineers,
store data. In the process of gathering evidence at a
entomologists, autopsy specialists, and odontolo-
crime scene, some equipment such as support tools
gists can work in the criminal investigation pro-
designed for specific purposes, mobile phone/ cam-
cess together or respectively [18], [19]. These team
eras/ computers, tele-forensic and telecommunica-
members work on evidence for making them visi-
tion technologies, hand tools, forensic light sources,
ble and collecting them from different places and
metal detectors, latent print kits, light sources and
surfaces. Also, it is the task of this expert group
tools, and software/hardware-assisted systems for
to protect the collected evidence and to conduct a
data enhancing and matching are used [11], [21].
preliminary examination [18], [19].
The investigation team works to collect the nec- The data with biological and biometric content
essary information from eyewitnesses, examines the from the collected evidence is transferred to the
immediate surroundings of the scene, and assist laboratory for detailed examination. Detailed labora-
the other teams and experts with the information tory examination also requires a series of procedures
they collect. The investigation team works to link and different equipment to perform these proce-
the crime scene data with each other and with the dures. With the infiltration of technology into all
incident and synthesize the collected data to create areas of life, it is not surprising that the techniques
a scenario that revives the incident [18]. and methods used to resolve crimes are futuristic.
From retinal scanning to evidence chemistry mon-
2.3 Technical Evaluation itoring, new forensic technologies can be far more
advanced than the methods used to solve crimes in
Assessing the crime scene requires an action plan. science fiction films. Nowadays, biometric recog-
After the scene is described and protected from nition systems using video and audio recordings,
external influences, it is depicted using sketching fingerprint identification using AFIS technologies,
or photography to evaluate the incident before spe- detection systems using DNA found in body fluids
cialized analysis begins. Subsequently, the visible such as saliva, sweat, and different technology-

93
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

Fig. 3: Most Common Forensic Disciplines [20]

supported designs are used in all in many CSIs.

3. Evaluation of Crime Scene Compo-

nents in Terms of Fingerprint

CSI requires a multidisciplinary study. While CSI

is a process that should be examined from a soci-
ological and psychological point of view in terms
of murderer-victim-cause, it can be considered as
a criminal case with the point of reconstruction of
the crime and going from evidence to the offender.
Therefore, the crime scene can be classified through
Fig. 4: Evaluation criteria of crime scene compo-
different perspectives considering the type/ effect
nents
of the crime, the status of the killer/ victim, the
type of data containing the evidence, the type of
methodology/ method planned to be used in the collection and analysis. In addition, each scene
analysis process, and the type of investigation. has its characteristics, and the characteristics of
Each type of data at the scene has different the incident and the environment should be taken
characteristics. These differences make it necessary into consideration during the investigation process.
to use different procedures in the process of data Due to the variations arising from the evidence,

94
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

environment, and incident, the approaches to be any coloring agent, it leaves a mark on the surface
proposed to improve and facilitate the CSI process thanks to the liquid secreted from the pores on it
should focus solely on solving a particular problem. [22]. This type of marks, which is not seen directly,
Within the scope of this study: is made visible by various physical and chemical
• Considering the scenarios in which different methods and used for culprit detection.
actions take place at different crime scenes, Plastic and patent fingerprints are easily visible
the difficulties encountered in the process of without the need for an external process. External
collecting and analyzing fingerprints from the components such as dust, sprays, reflective light
scene were identified and these difficulties were sources, or chemicals are used to detect and make
dealt with a computer science perspective. latent fingerprints visible [23], [77]. If the finger-
• The CSI process was divided into components: print is detected on the surface of a portable object,
the acquisition of fingerprint evidence, the ex- it is collected as evidence and examined in the
perts involved in the collection of the finger- laboratory. If fingerprints are detected on surfaces
print, and the management of the technical that are unlikely to be moved, such as doors or
process. walls, the data collection procedure is terminated by
• In order to ensure an objective evaluation of viewing these marks with standardized apparatuses
the process steps performed by each component for digital imaging (cameras, scanners), or using
and during the interaction of the components, different methods such as taping.
criteria were developed in light of the previous The photographing of mark evaluated within the
studies in literature, as seen in Figure 4. scope of the study is called technical photographing.
Patent fingerprints can be photographed directly.
3.1 Fingerprint Detection at Crime Scene Latent fingerprints are made visible by using special
filters, lighting techniques or chemical enhance-
The methods that are planned to be used in the
ments and recorded with technical photographic
process of detecting and collecting the fingerprint
method [4], [21], [24].
are selected by considering the characteristics of
the surface and environment where the fingerprint 3.2 Experts Working in Fingerprint Collection
is located. In this case, a general classification of Process
fingerprints is mandatory. Fingerprints are basically
classified into three main groups as patent, latent, Police officers and detectives do not have the spe-
and plastic [22]. A patent fingerprint is formed cial skills and expertise needed to assess all criminal
by transferring a material containing color such incidents. For this reason, support staff specialized
as blood, paint, and dirt from the finger to the in various fields can produce results by using their
surface. Such marks can be seen with the naked eye skills in CSI. CSI officers and evidence technicians
without the need for external assistance. The plastic are specially trained police or civilian personnel for
fingerprint is a three-dimensional pattern on a soft scene processing [11]. They are equipped to collect
surface. These marks are collected from surfaces and store physical evidence at the scene. Experts can
such as soap, mud, wax, paste. Latent fingerprints gather evidence at the crime scene by using a variety
are patterns that are not easily visible to the naked of tools and equipment, thanks to the training they
eye. Even if the finger touches a soft surface without receive [11].

95
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

Different disciplines work on fingerprints at the ment, removing the characteristic feature from the
scene. They are classified as these who detect finger- fingerprints are evaluated, and the software/ hard-
print data, who photograph the data, and who trans- ware components and technological infrastructures
fer the data for detailed examination. Identification used in comparing the fingerprints collected with
officers (ID) working at the scene are responsible for the registered samples are examined.
detecting latent fingerprints [11]. Fingerprint ana-
lysts, often referred to as latent analysts, are those
3.3.1 Fingerprint Collection Methods
who are responsible for maintaining, inspecting and
evaluating fingerprints, which are part of the CSI.
For this reason, experts who record the scene with There are established standards for devices used
photos or videos are required to receive training for imaging in CSI. These devices range from mid-
to operate and maintain the required sophisticated level professional devices to high-level professional
equipment, and to produce high-quality records with devices, depending on the budget of the investiga-
appropriate scales by protecting data integrity [24]– tion department and the specificity of the incident
[26]. Biometric evidence such as fingerprints, palm [4], [24].
prints, footprints collected at the scene accelerates In order to detect and make latent fingerprints
the detection of the offender and helps to conclude visible, many procedures are applied. In these pro-
the case directly. For this reason, specialized per- cedures, fingerprints are made visible by using com-
sonnel who work in CSI teams should show very ponents such as external powders, sprays, reflective
high sensitivity when collecting such biometric or light sources or chemical substances [23], [77].
biological data. This sensitivity extends the CSI Once the fingerprint becomes visible, it is stored
process and requires multiple experts to work si- for detailed examination by video recording or
multaneously [11], [27]. photographing.

3.3 Technological Infrastructures Used for Col-

lection and Evaluation of Fingerprint 3.3.2 Fingerprint Processing

While technological advances affect every aspect Fingerprints collected from the crime scene are
of life, their impact on the CSI process is not not of the same quality as fingerprints collected with
surprising. One of the factors that directly affect the devices such as a sensor or scanner. Due to the am-
success of fingerprint experts is the equipment used bient conditions, the fingerprint data may not have
in data collection and evaluation procedures. As the sufficient characteristics, only a part of fingerprint
crime scene becomes more complex, the number may be collected instead of the whole finger, the fin-
of equipment used increases, and more specialized gerprint may not have sufficient reference point due
products are needed. to the noise in the environment. In addition, clearing
In this section, the procedures applied to fin- the noise in the received fingerprint data, separating
gerprints in the CSI steps which are making the the fingerprint from the scene, and the removal of
fingerprint visible, transferring the fingerprint to the incorrect reference points in the fingerprint make it
laboratory environment for detailed examination, easier to determine who the fingerprint belongs to
analyzing the fingerprint in the laboratory environ- because these processes increase the match score.

96
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

used, fingerprints are classified according to their

physiological characteristics to enable one to many
searches. Catalogs of fingerprints are grouped with
a classification method based on Henry’s system,
and matching of each fingerprint was performed
under expert control. Because of the increase in
population, the process of pairing the fingerprint
with the traces in the database led to an increase
Fig. 5: Improving the fingerprint collected from the in the number of comparisons. This increase leads
crime scene to another increase in workload and time spent,
and human resources used [31]. In the criminal
justice system, fingerprint recognition systems have
The collected data is not used directly. Before
been used for more than a century to detect suspect
removing the characteristics of the fingerprint col-
in crime and terrorism in order to support public
lected from the scene, the determination of the
security. During the criminal investigation, the latent
singular point, calculation of the orientation and fre-
fingerprints, which are found, developed and im-
quency of the ridges, segmentation, normalization,
proved at a crime scene, are compared with the pre-
filtering, thinning processes are performed [28],
viously registered fingerprints of the persons. The
[29]. A fingerprint in Figure 5(a) taken from the
fingerprints left behind at the scene, whether full or
scene changes to Figure 5(b) after the enhancement
partial fingerprints, can be identified with today’s
steps applied. There are many academic studies on
technology. A list of potential candidates similar to
fingerprint improvement. The studies on fingerprint
the fingerprint to be identified can be found using
image enhancement in the literature have been sub-
image recognition algorithms. The systems that do
jected to a general evaluation, and the results are
this automatically are called AFIS. For this reason,
summarized in our previous study [30].
AFIS has been seen as a tremendous improvement
for local, state and federal law enforcement agencies
3.3.3 Fingerprint Matching and Storing to manage fingerprints and to identify people [31].

As a result of the research conducted to determine 4. Problems in Forensic Fingerprint

which characteristics should be emphasized in the
Analysis
detection of the similarity of the fingerprint, features
called minutiae were discovered for the fingerprint The basic principle of CSI is to obtain material
specialists to use in manual examinations. Using findings, to separate legal evidence from the other
these features, which consist of different positioning
collected findings, and to use legal evidence for
of the fingerprint ridges, the topological equivalence
arresting the perpetrator. In the CSI, the outputs ob-
of the data from the two fingerprints can be analyzed
tained from the evidence collected are evaluated and
[31]. After the completion of the preprocessing criminal action is taken against the perpetrator. For
step, the fingerprint collected from the crime scene
this reason, even the slightest omission made during
becomes suitable with the minutiae extraction. the detection, classification, collection, laboratory
Before automated fingerprinting technologies are examination of findings may result in significant

97
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

Fig. 6: Problems Encountered in Fingerprint Collection and Analysis

errors in the judicial process [74]. incident should be examined and in which aspect
Subjectivity is one of the main problems that the crime scene components should be evaluated.
concern many areas of forensic sciences and fin- The environment in which the incident takes place
gerprint analysis. The CSI process is shaped by determines which specialist, equipment, method,
tested scientific methods and the analyst’s subjective methodology, and technique should be used. For
inferences. The hypothesis of how the incident took this reason, possible scenarios in which potential
place depends on the examination of the analyst and problems may occur were developed in this study
interpretation of the evidence. Another key problem in order to illustrate how the components of the
is the inability to undo any action taken during investigation were affected by the crime scene.
the review process. If the procedure to be followed If the necessary precautions are not taken to
is incorrect, a non-reversible path has been taken. protect the environment after the first arrival to
These two problems form the basis of every problem the scene, the whole analysis process is in danger.
that can be encountered during the CSI process. Fingerprint specialists are forced to perform more
In this part of the study, the problems encountered complex processing steps in crime scenes that are
in CSI (Figure 6) are discussed. These problems not properly protected. Experts may have to use
have been explained under a number of questions. more specific methods and equipment during the
data collection process in these scenes. In addition,
4.1 What are the types of crime scene? when the collected data becomes contaminated,
external processing steps should be used in the
The environment in which the crime was commit- detailed analysis of the collected data. In this case,
ted is considered a key point in terms of how the the collection of fingerprint data can become time-

98
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

consuming and costly. In addition, integrity and of the specialist in the use of the equipment can
validity may be violated due to the corruption of guide the scene investigation process. For this rea-
fingerprint data [2]. son, to illustrate how the selection of equipment to
If the crime scene has a chaotic structure due to be used and procedures to be applied to fingerprint
the way the crime is committed, critical decisions collection affect the components of CSI, problems
should be made on both the working specialist and were given in the following possible scenarios.
the equipment used. If the number of fingerprints Crime scene investigators can use many differ-
that need to be examined increases, more specialists ent methods to photograph an incident. Nowadays,
should work at the scene. However, due to the video recording and digital photographing methods
inadequacy of staff with the appropriate criteria to are widely preferred due to their ease of use com-
work in the field, the investigation is usually carried pared to film rolls and videotape technologies. It
out with a small number of specialists. In this case, should be noted that the devices used should have
the collection of fingerprint data can become time- special features required for CSI and should com-
consuming and costly [33], [34]. ply with standards accepted by forensic sciences.
The mistake in terms of experts worked and Parameters such as memory cards, lenses, filters,
equipment used in the fingerprint collection process external light sources, reflective chemicals, and the
creates irreversible effects. Data collection may fail age of the devices constitute a reliability problem for
if the expert does not have sufficient field knowledge the collected evidence. Keeping technical equipment
or does not have the experience and ability to in extremely hot and humid environments, lack
overcome factors arising from the nature of the of maintenance, failure of bringing extra storage
environment [35]. units and external batteries make the investigation
of the scene difficult [37]–[39]. That’s why data
All techniques and technological infrastructures
integrity and validity are at risk if the procedures
used in the visibility, storage, improvement, and
and equipment used in the fingerprint collection
matching of the fingerprint must be legally valid.
process do not comply with standards accepted by
Interventions at these stages should not affect the
forensic science.
uniqueness of the collected fingerprint data. If these
steps are not designed appropriately, it is inevitable The fact that the devices used for CSI have the
that there will be problems in terms of cost, validity, appropriate criteria does not eliminate the whole
and integrity [36]. problem. In addition, the technical knowledge of the
expert using this device directly affects the process.
4.2 What are the equipment requirements and In particular, evidence may not be considered legally
collection procedures for fingerprint recording? valid if the analyst monitors it without taking into
account certain criteria such as fingerprints, correct
angle, light, and resolution detected during forensic
The selection of equipment used to collect finger-
photographing [37]–[39].
prints from the crime scene and procedures to be
followed at the scene is a critical process. Consid- In order to complete the collection of finger-
ering the environment in which the incident took print data in the most trouble-free way, innovative
place, the suitability, performance, usability and va- approaches are needed in terms of experts and
lidity of the equipment, as well as the performance equipment used in the examination of the incident.

99
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

As technology advances rapidly, the functionality the wrong place and analyzing the incorrect data
of all types of equipment has increased and can increases the cost as well as causing the loss of time,
offer higher performance. However, features added adversely affecting the performance of the expert,
to improve performance can increase the complex- and leading to misuse of resources [11], [18].
ity of the equipment. Because of this complexity,
specialists may need additional training to use this 4.4 What are the collection methods of the
equipment, and it can be seen as a cost disadvantage. fingerprint from different surfaces?
Although the increase in functionality provides an
advantage in terms of validity and time in the data The methods of collecting the fingerprint vary
collection process, it can be disadvantageous in according to the surface on which it is located. On
terms of complexity and cost. the solid surface, fingerprints are made visible by
dusting and are collected by taping or photograph-
4.3 Where to look for fingerprints? ing, while on soft surfaces fingerprints are collected
by photographing. The fingerprints of the suspect
In a theft investigation, the perpetrator’s entry may be detected on clean, dry, hairless parts of
point should be where the fingerprint model begins the victim’s body, and these marks are likely to
to be searched. The door handles can be the first be recovered. To collect this kind of fingerprints,
points to look at, or if the burglar has entered a variety of chemical, electronic and photographic
through the window, it is possible to have finger- techniques should be applied, as well as dusting.
prints on the window handle or glass. In the case of As can be seen from these examples, a different
an injury or death, the murder weapon at the scene, procedure is required for each surface and each
the surfaces likely to come into contact during the procedure requires different expertise and forensic
fight, or the fingerprints left by the killer on the technology. Experts can try to collect fingerprints by
victim should be handled. If there is access to the using incorrect methods and components if they do
scene after a long period of time, the environment not have sufficient field knowledge. It is a serious
may be polluted by living things such as animals, problem that it is not possible to correct the dete-
people, or for natural reasons such as rain and wind. rioration in the fingerprint that is attempted to be
In such cases, it is a very technical detail to decide collected with these wrong choices. Such situations
which region the fingerprint should be searched and pose a great risk to the validity of fingerprint data
by which techniques. If it is necessary to investigate collected from the scene [37], [40].
a crime scene in such specific cases, it is essential
to collect prints that are directly relevant to the 4.5 How to make task assignments in CSI?
incident by the appropriate technical infrastructure
and experienced investigation personnel. The faster Crime scene investigators are responsible for eval-
solution is generated, the easier it is to catch the uating evidence at crime scenes, photographing/
perpetrator. packaging/ transporting evidence and creating re-
However, if the process of analysis and produc- ports and testifying in court. Investigators should
ing the results is prolonged, those who commit have a clearly defined role in the tasks in which
the crime or participate in the crime may escape they are adequately trained to perform. In some
or disappear. Also, searching for the evidence in crime scenes, staff may need to play more than one

100
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

role because of the size and type of the scene, or Certification programs are not particularly preferred
because the incident has to be evaluated with fewer by law enforcement officers in terms of cost and
experts than it needs. Improper staff appointments time, especially when considered in terms of the
cause the evidence to be collected incorrectly. This participation of staff working in small areas [41].
situation causes major mistakes in the investigation
process [37], [41]. In addition, the sudden increase 4.7 What are the key attributes of personnel?
in the number of incidents per day or the specific
circumstances in which a series of crimes are con- By using international standards and procedures,
sidered as serial murders may directly affect the existing problems in the data acquisition and tech-
number of officers and specialists working within nical evaluation components can be overcome, and
a day. Unsuitable assignments for personnel cause improvements in the results obtained from these
an excessive workload. The fact that non-specialist components can be made. However, the field ex-
staff is employed at the scene or the expert staff is pertise component, the subject of which is human,
exposed to working hours exceeding their capacity should be evaluated from different perspectives. The
invites mistakes in the CSI process [41]. Especially physical and mental characteristics of the staff affect
when the fingerprint collection from the scene is a the success of the CSI process. The expert who
complicated process, non-specialist personnel may will work at the scene should be hired considering
disrupt the integrity of the crime scene and the the parameters such as cognitive ability, experi-
chain of custody. In cases when a small number ence, physical trait, mental endurance. However,
of specialist personnel is overloaded, the cost of as experts having these criteria can be costly for
the CSI may increase due to the overtime of the law enforcement agencies, the units firstly recruit
specialist, and there may be a risk of evidence civilian personnel for short-term employment for
invalidity due to the possibility of a loss of quality economic reasons. With this method, it is aimed to
in the collected data. solve the need of experts for less cost temporarily.
This situation is a major chaotic problem as experts
4.6 What is the specific training for CSI? cannot gain experience at different crime scenes
[35].
There are three different crime scene examina- In all law enforcement agencies, the staff working
tion certificates. These are Certified Crime Scene in the crime scene unit are uniquely stressed com-
Investigator (CCSI), Certified Crime Scene Analyst pared to other officers. Because they are closest to
(CCSA) and Certified Senior Crime Scene Analyst the crime scene and spend quite a long time at the
CSCSA) [42]. They are internationally recognized, crime scene. Also, CSI is stressful because critical
and in order to receive them, a high level of ex- judicial decisions are made based on the results of
perience and training is required. With these train- this process, and these decisions vary according to
ing contents, crime scene investigators can become the success of the investigation. Failure to provide
qualified personnel. The most crucial problem is the a sufficient number of experts due to criteria such
difficulty of law enforcement agencies to provide the as cost, time and experience during the CSI process
budget for these certificate programs and whether causes the team to be subjected to heavy working
the person who wants to follow these programs tempo. Studies have shown that even experienced
can find the external time to attend the courses. personnel still have the same level of stress in CSI.

101
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

In addition, personal characteristics directly affect proposed methods on data repair/ restructuring are
stress levels. As with any profession, extreme stress considered legally invalid in law enforcement. In
leads to errors, and this inference applies to CSI particular, there is a lack of alternative approaches
teams. Excessive stress brings risk. For this reason, that can be used when performing partial finger-
it is necessary to adopt new approaches to alleviate prints analysis, if there is not enough reference point
the workload per employee [43]. [33], [34], [48], [49].

4.8 What are the standards of forensic science 4.10 How much information is required for
technologies? identification/verification?

Under normal circumstances, there is only one Even if the fingerprint data was collected through
opportunity to conduct an investigation at the scene the most appropriate procedure and improved by the
and to collect the relevant material from there. For best preprocessing methods, the acquired properties
this reason, each step must be carried out metic- may not be sufficient to identify the person. The
ulously. Complete registration of all areas where features extracted from the fingerprint may not be
fingerprints are made visible should be done using specific enough to identify the person or sufficient
appropriate methodologies and equipment which features may not be extracted when the fingerprint
are accepted as valid by judicial authorities. The is partial. When the feature group extracted from
material used to display all evidence at the scene the fingerprint is not distinctive enough to identify
is determined by international ISO/IEC 17020 and the person, many matching examples can be en-
ISO/IEC 17025 standards [44], [45]. In addition to countered in the AFIS query. If there is not enough
the standards mentioned above, all technologically feature for evaluating the fingerprint, it is necessary
assisted data collection devices may adversely affect to examine the fingerprints which are filtered by
the validity of the collected data [46], [47]. using AFIS in detail by fingerprint experts and
determine which pattern matches precisely. In the
4.9 What is the relationship between image absence of sufficient characteristic points in partial
processing and validity? fingerprints, the collected fingerprint data is not con-
sidered valid. These situations reveal the need for
Despite advances in image processing, the ex-
new approaches in fingerprint matching. Alternative
amination of the latent fingerprint is complicated
perspectives should be developed in cases where
process. It is a complex task that is mainly depen-
the fingerprint is partial or the characteristics are
dent on the visual judgment of highly educated hu-
insufficient [34], [50]–[52].
man supervisors. Fingerprints collected from crime
scenes contain less information than fingerprints 4.11 How reliable are fingerprints in solving
collected under controlled conditions. They are of- crimes?
ten noisy and distorted and may contain a partial of
the total fingerprint pattern. It is known that the less In the 2010 report issued by the US National Insti-
the intervention to the fingerprint collected from the tute of Justice, it was noted that automated systems
scene, the higher the validity of the data. For this made less accurate conclusions than well-trained
reason, approaches that lead to loss of character- Fingerprint examiners while comparing latent fin-
istic features in the data improvement process or gerprint collected from the incident with fingerprints

102
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

registered from the database. On the other hand, technical infrastructure, especially in local areas,
the US National Research Council’s 2009 report old technology-supported approaches continue to be
on forensic science is a milestone for fingerprint used.
biometry. This report states that the only forensic
method that has been rigorously verified is nuclear 4.13 How complex is the technical equipment?
DNA analysis and that all other forensic sciences,
including fingerprints, are questionable. Due to high The most desirable approach in the CSI process
mismatch rates and expert requirements, the current is that the fingerprint, which is easily visible to the
fingerprint matching methods have been questioned field expert, can be collected effectively and pro-
in this study and a need for external approaches cessed in the simplest methods to produce a result.
for fingerprint identification was detected. Today’s The increasing complexity of the techniques and
AFIS is known to make more errors, especially in methods used to manage crime scene components
partial fingerprints, because it uses the minutiae- consequently leads to an increase in time spent and
based matching method. It has been observed that cost. The more complex the equipment used, the
new approaches should be developed apart from lower the number of experts who use the equipment
the similarity of minutiae points. However, if all effectively, and this reduces the performance of the
software and hardware supported mechanisms used CSI [49].
in fingerprint recognition need to be rearranged for
a new approach to be proposed, a major transforma- 4.14 Why is teamwork important?
tion process in terms of applicability, cost and time
may need to take place [36], [75]. The most critical element in CSI is teamwork. The
investigation of criminal proceedings is usually car-
4.12 Technical Equipment Cost ried out by the joint evaluation of persons working
in different institutions. This system is designed in
Technical equipment used to make the fingerprint such a way that no person or institution can work
visible (chemicals/ powders/ light sources), tech- independently. For this reason, the multidisciplinary
nical equipment used to collect fingerprint (video working process involves problems such as defining
recorders, cameras, high-resolution lenses, scan- to whom the responsibility belongs and who should
ners), software/ hardware-assisted equipment to pro- perform which task. If appropriate authorization
cess the fingerprint (Image enhancement/ recon- is not made, CSI can become a chaotic process.
struction software, 2D-3D image converter soft- Crime-management approaches designed far from
ware), the infrastructure used for matching and a result-oriented approach may have poor results in
storing (AFIS, IAFIS, ALFIS, MAFIS) constitute terms of time, cost, integrity, and validity [2], [11],
the technical evaluation component of the CSI [53], [18].
[54]. Depending on the capacity of the country,
the requirements such as the supply/ installation/ 5. Challenges in Forensic Fingerprint
maintenance/ repair of the infrastructures needed to Analysis
store, protect and analyze millions or even billions
of fingerprints are cost-increasing factors [47], [52], In order to accept the fingerprint as evidence by
[54], [55]. However, due to the lack of sufficient the forensic sciences, three essential claims need to

103
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

be confirmed: (a) each individual has a unique and cient by itself without consulting an external
permanent fingerprint, (b) fingerprint collected from expert in court.
the scene can be used to identify the individual,
For this reason, in order to achieve the afore-
(c) the fingerprint recognition system has to have
mentioned objectives and to eliminate the problems
zero or near-zero error [76]. However, the results
mentioned in fourth section or to minimize the
obtained from software and hardware supported in-
effects of these in the forensic fingerprint exami-
frastructures to verify these claims are not sufficient
nation process, the difficulties to be overcome are
enough. Because:
examined in detail in this section.
• the fingerprints of individuals can be very sim-
ilar, 5.1 Improvement of Fingerprint Identification
• the collected fingerprint may not have many Methods
characteristics,
• as a result of a query in AFIS, there may Most of the fingerprint recognition/ verification
be multiple samples that are similar to the mechanisms are based on the matching of minutiae
fingerprint collected. points. Minutiae point-based techniques analyze fin-
Thus, software and hardware supported mecha- gerprints according to local characteristics such as
nisms only minimize the number of suspects. Then, termination and bifurcation. If the matching scores
fingerprint experts determine which pattern exactly of the Minutia points are above the threshold level
matches with each other. Forensic fingerprint anal- set by the system, the two fingerprints are matched.
ysis has been used for over a century as a way of This approach has been extensively studied and is
identifying criminals. In order for the fingerprint to the backbone of most of the current AFISs [73].
be considered as evidence, analysts have to claim Fingerprints collected by devices that are used to
absolute certainty that the fingerprint has been left digitize fingerprints (sensors, scanners) or by tech-
by a suspect. In summary, the decision made about niques such as the inking method used in forensic
fingerprint identification is based on an expert’s anthropology have many minutiae points. When an
opinion instead of scientific data. AFIS scan is performed using these patterns, it is
effortless to determine who the fingerprint belongs
When all of the problems encountered in crime
to. Since fingerprints in AFIS databases are col-
scene analysis were evaluated on the basis of en-
lected for use in biometric recognition/ verification
vironment, people, methods, and the headings in
mechanisms, the collection of fingerprint data is
fourth section were determined. Thus, the main
performed in environments where there is no data
goals in CSI are:
loss. However, to express the same situation for
• To optimize the scene for the investigation, invisible fingerprints is quite out of the question.
• To have the opportunity to maximize data qual- Fingerprints collected from the scene contain many
ity by having every evidence gathered by the problems. First, the fingerprint may have been ex-
most experienced experts, posed to noise. In addition, the fingerprint may
• To ensure that the methods used to collect and be partial and the collected part may not have
analyze evidence are legally valid, sufficient characteristics to identify the person. As a
• To ensure that the evidence collected is suffi- result of a comparison between the latent fingerprint

104
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

collected from the scene and those in the AFIS conducts all stages of the techniques and methods of
database, it is possible to obtain many matching data collection. However, a crime scene can some-
fingerprints. For this reason, different approaches times have hundreds of invisible fingerprints. This
should be considered in order to examine especially results in an increase in the time of the CSI and a
low quality or partial fingerprints. The first option delay in the conclusion. That’s why a more practical
is to find another distinctive criterion that can be method for data collection from the crime scene
used instead of minutiae for fingerprint recognition. needs to be modeled and it is thought that such a
One of them is the use of Level-3 characteristics model will accelerate the process. To achieve this,
which fingerprint analysts use to separate similar approaches to fingerprint collection with equipment
fingerprints. that does not require professionalism are needed
There is no general definition of Level 3 char- to accelerate data collection procedures. Scenarios
acteristics; they can be everything such as scars, capable of expressing, defining and making sense
deformations, pores, and ridge shapes. Since even of the data acquisition stage should be developed
a small portion of the fingerprint contains many in order to indicate in which environment and by
of these features, it is considered to be suitable which method the fingerprint data will be collected
for evaluating fingerprint data collected from the independently from the expert. Exemplary proto-
scene. However, the biggest problem in identifying types such as Forenscope, Wampire, Papilon Fosko
the fingerprint using Level-3 features is that the fin- have been developed to be used in this context
gerprint data used must have a high resolution. The [56]–[58]. In order for these prototypes to develop
fingerprint data collected by using low-resolution evidence collection capabilities, different disciplines
sensors are not suitable for this method. When should conduct joint studies in order to collect valid
the fingerprints registered in the AFIS databases fingerprint data.
are evaluated within this scope, no matter how
many characteristics are removed from the partial 5.3 Improvement of Data Storage Infrastruc-
fingerprint data, it is not possible to use the Level- tures
3 characteristic features because fingerprint data in
AFIS has not been collected by considering Level-3 Detection of a person through a biometric fea-
characteristics and it is not possible to use level-3 ture is important for many different applications in
characteristics in a scenario where the same data criminology and forensic identification. The finger-
cannot be re-collected [59]. print is seen as an effective tool for identifying
the individual because of universality, uniqueness,
In order to work with fingerprints in the existing
and permanence. Many different approaches have
databases and to improve the low matching rate due
been proposed in the literature to accurately deter-
to insufficient minutiae points, deep learning-based
mine whether two fingerprints belong to the same
approaches are seen as a solution [60].
person. Among these, matching mechanisms based
5.2 Development of Automatic Evidence Col- on minutiae points stand out as the most accurate
lection System techniques due to their distinctive capabilities. How-
ever, fingerprint recognition systems running with
Making the latent fingerprint visible is a time- a massive database have some difficulties such as
consuming process, and a crime scene examiner lack of scalability, low performance and high cost of

105
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

complex matching algorithms, and loss of accuracy institutions and adversely affects the adoption of big
due to the comparison of millions of data. data technologies. However, the increased data size
Large-scale automated biometric recognition sys- requires the use of such technologies.
tems have many specific requirements:
5.4 Development of Evidence Association
• The system should provide reliable identifica- Models
tion services with large databases.
• The system should be able to support existing Fingerprints are fully formed in the womb dur-
and evolving standards. ing the nine-month development of the baby, and
• System-generated templates or databases must ridge-valley configurations in the fingerprint do not
be suitable for use on different platforms. change during an individual’s life unless an external
• If clients are remote from the server hosting the condition such as an accident or injury happens [61].
central database, the system must be able to run Although biometric features such as gender and
on a network. age do not have distinctive features to distinguish
• Latent fingerprints must be processed before the individual, they are used as additional evidence
being sent to the system. to complement primary biometric identifiers such
• The cost to design a system with these require- as fingerprints, face, and iris [62]. During crim-
ments must be acceptable. inal analysis, biometric data such as fingerprints
There is a need for infrastructures, techniques and blood collected from the crime scene, records
and technologies that can produce fast and error- of the security camera at the time of the crime,
free results with fingerprints and have no problems statements of eyewitnesses are analyzed and cor-
in analyzing and processing ever-increasing data. related with each other. In light of this information,
Biometric systems based on big data technologies the reconstruction model of the crime is formed.
should be considered as an alternative approach Approaches such as the assessment of whether the
to meet similar requirements and to contribute to blood and fingerprint taken from the scene belong to
the solution of different problems in CSI. In a the same person, and the association of gender and
recent report published by Forbes, it was stated that age information collected from security cameras/
Big Data Technologies are being used to establish, eyewitnesses with fingerprint patterns directly affect
manage, and maintain the register of citizens by the criminal investigation process. Such approaches
UIDAI [61]. The biometric research group defines can be used for selection (or elimination) between
big data as ”a term used to describe big and complex matching samples, especially where the invisible
data sets that can give a clearer result when analyzed fingerprint matches a number of samples in the
meaningfully” [62]. database records.

Big Data technology is expected to be used by law Table 1 presents the studies in which fingerprint
enforcement officers for evidence association and is associated with biometric and biological char-
identification, as it incorporates a variety of data that acteristics. If the correlation between fingerprint
provides the opportunity to relate different biomet- characteristics, facial features, blood type, gender
ric characteristics of individuals as well as finger- and age information are to be evaluated:
printing. Renewing or completely replacing existing • the relationship between the fingerprint and
technological infrastructures creates high cost for biological/ biometric data of the individual is

106
 Ref. Associated Main Purpose Methods Dataset Success Rate
Feature
[65] Fingerprint/ To identify the dominant fin- the t-test in SPSS and p<0.05 The trace of ten fingers In women, 55.28% of the ulnar structure
Gender gerprint patterns of the South of 250 men and 250 26.84% whorl/ in males, 49.32% ulnar 30.64%
Indian population and the prob- men whorl structure was observed
ability of gender distribution
between fingerprints
[66] Fingerprint/ To predict gender based on the Bayes theorem 275 men and 275 Men’ ridge density is less than 13 ridge / 25
Gender density of fingerprint ridges women between the mm2 and women’ ridge density is larger than
ages of 18-65 14 ridge/ 25 mm2
Arslan et al., Vol.8, No.4, pp.88-113

[67] Fingerprint/ Estimation of sex using ridge Fuzzy CMeans (FCM)/Linear 10-fingerprint images Obtained results of 80.39%, 86.5%, and 88.5%
Gender count, ridge thickness to valley Discriminant Analysis (LDA)/ for 2200 persons of using FCM, LDA, and NN, respectively
thickness ratio (RTVTR), white Neural Network (NN) different ages and
lines count, ridge count asym- gender (1100 males
metry, and pattern type concor- and 1100 females)
dance
[68] Fingerprint/ To generate the face border of Artificial Neural Networks 120 people Fingerprints and face borders have relations
Face a person using only fingerprint among each other closely
biometric feature of the same
person without any information
about his/her face
[69] Fingerprint/ To analyze the existence of any Artificial Neural Networks 120 people Developed system generates the stationary face
Face relationship among fingerprints parts of a person including eyebrows, eyes and
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE

and face parts nose from only one fingerprint image of the
same person without knowing any information
about his or her face with the errors among
1.4% and 4.8%
[70] Fingerprint/ A quick and efficient technique Neural Network Classifier for The approach is tested After testing the fingerprint images with the
Blood Group of blood group detection using feature extraction over a dataset of 4 training set of fingerprint images and the
a set of features of the finger- images in each blood approximate result of 80% of matching is
prints group achieved
[71] Fingerprint/ To develop an algorithm for 2D- Discrete Wavelet 300 Fingerprints be- The age information of the sample fingerprints
Age estimating the age through fin- Transform (DWT)/ Principal longing to the various tested is shown. No external success rate has
gerprint Component Analysis (PCA)/ ages in between 1 – 80 been given
The Support Vector Machine
(SVM)
[72] Fingerprint/ To develop a procedure to ex- Extracted features into 250 images for train- The results are shown in tables and graphs
TABLE 1: Studies on the association of fingerprint and other biometric properties

Age tract discriminating features us- the principal component ing and 100 images for
ing Curvelet Transform to clas- (PCA) subspace/ Age group testing
sify fingerprints into three age estimation using K-nearest
groups neighbor (KNN)

107
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

evaluated with two approaches. In the studies ferences between those proposed in theory for
examined, it is stated that fingerprint is related scene investigation and those applied in prac-
to blood group, age, gender, and face shape. tice. CSI process models, proposed procedures
The relationship between the fingerprint and for data collection, techniques and technolo-
the other features mentioned is evaluated sta- gies, and data processing methods have been
tistically and the distributions of which fea- designed taking into account the case exam-
ture or feature combinations of the fingerprint ples that have been experienced so far. How-
are related to blood group/ gender/ age infor- ever, in the case of an incident scene that
mation and facial features is calculated. The has not been observed so far, the experience
other approach is about deriving the relation- and knowledge of the law enforcement forces
ship model between fingerprint characteristics that manage and illuminate the process, the
and blood group/ gender/ age information and expert’s innovative approach, experience, and
facial features and estimating other biometric analytical thinking ability, the used equipment’s
and biological characteristics of an individual modification capability, diversity, capacity, and
by using a fingerprint. functionality criteria play a critical role. For
• statistical or learning-based approaches are used this reason, the most important and risky part
to estimate other biometric properties from fin- of the CSI is the approaches to identify the
gerprints. crime scene and data collection, the disciplines
• the relationship between different biometric to be studied, and the stages of determining
properties can be used to understand the rela- the technical infrastructure. Also, to minimize
tionships between evidence collected from the the problems encountered at these stages, data
scene. In this way, the data that should be collection/ improvement/ analysis mechanisms
investigated in the chain of custody can be and procedures to be applied by the expert
determined. should be determined and automated from a
• priority analysis of data directly associated with systematic point of view.
the fingerprint at the scene can be carried out. • Cognitive and Human Factors: No matter how
If there are fingerprints that are not related automated the CSI process is, it is impossible
to the incident but are somehow detected in to produce a definite result without the human
the environment, priority may be given to the factor with today’s theoretical and practical
examination of the directly associated data. In knowledge. In the CSI, where the intervention
addition, disregarding non-relevant data will re- of a field expert is considered to be mandatory,
sult in significant time and cost savings. the most appropriate action to improve the
process and produce the perfect result is the de-
6. Recommendations velopment of technological software/ hardware,
tools and mechanisms to reduce the workload of
Within the scope of the information and findings the human factor. In addition, training facilities
obtained, the study was evaluated under different should be provided to improve the expert’s field
titles. The summary of this evaluation is given under knowledge. Finally, the criteria that will ensure
the following sub-headings: professional competence should be determined
• Crime Scene Realities: There are great dif- in the most appropriate manner and personnel

108
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

should be recruited according to these criteria. the work.

Forensic science often involves interpretation • Future Approaches in Analytical Crime Sci-
and subjectivity. Therefore, cognitive and hu- ences: From the perspective of 2025, forensic
man factors are important for making the most sciences will take a new direction with break-
accurate judicial decisions in critical processes. throughs in the fields of large-scale data anal-
To maximize objectivity and decision quality, ysis, self-learning mechanisms, new analytical
forensic researchers should receive cognitive techniques, and digital forensics. Undoubtedly,
bias training and evaluate only contextual in- technology will play an important role in the
formation. future of forensic science and how it is used
• Identification: The software and hardware- in both legal and criminal practices. One of
based mechanisms currently used in fingerprint the cornerstones of social life is the proper
matching should be updated periodically due realization of justice. The slightest mistake can
to the impact of new approaches and inven- cause an individual’s life to be shaken fun-
tions. Forensic experts, law enforcement and damentally, and the consequences may cause
judicial decision-makers are therefore inevitably reactions while deeply harming society. The
strengthened by a strategic working alliance correct decision of the judiciary depends on
with their industry counterparts. the existence of a large number of correctly
• Standards, Validation and Accreditation: collected and evaluated evidence. In addition
Standards have been developed to help practi- to the known and applied classical methods,
tioners use technology and methods in a clear information from other disciplines is needed
and robust way. The use of equipment and today. Also, it is possible to intervene in the
procedures in accordance with the standards new crimes and criminals that have emerged
will also protect the validity of the collected with the development of technology by the in-
data. In cases where the process is evaluated formation provided by new disciplines. Society,
in accordance with standards and there is no science, and technology are dynamic structures.
violation of the rules arising from the expert, It is necessary to develop new solutions to new
the CSI process is accredited and the validity situations that arise. For this reason, all forensic
of the hypotheses presented legally increases. sciences used in the CSI have to keep up with
For this reason, all the steps of the crime scene this development.
investigation should be designed in accordance
with the standards. 7. Conclusions
• Reliability of Forensic Sciences: Forensic sci-
ence plays an important role in the effectiveness The latest advances in biometric technology,
of the Criminal Justice System. To retain confi- equipped with computational intelligence tech-
dence in that forensic approach, it is important niques, replace manual identification approaches
to consider how they are implemented; not only in modern forensic science and fill the gaps in
in terms of its scientific accuracy, but also to the systems. In this study, the contributions and
show that evidence is handled appropriately to limitations of biometrics in the field of forensic
avoid contamination, and to demonstrate the im- identification are explained. The scope of the study
partiality and integrity of the staff undertaking can be expressed as:

109
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

• The examination of the fingerprints at the scene References

was focused on and the elements necessary for
[1] H.C. Lee, E.M. Pagliaro. ”Forensic evidence and crime scene
the successful realization of this process were
investigation,” Journal of Forensic Investigation, 1(2), pp. 1-5,
identified. 2013.
• The scene investigation process was divided [2] J. Reno, D. Marcus, L. Robinson, N. Brennan, J. Travis. ”Crime
Scene Investigation A Guide for Law Enforcement,” 2000.
into three components: information acquisition,
Accessed on: Dec. 3, 2019. [Online]. Available: https://www
field expert activity, and technical evaluation. .ncjrs.gov/pdffiles1/nij/178280.pdf.
• The relationship between each component and [3] R. Gehl, D. Plecas. ”Introduction to Criminal Investigation:
Processes, Practices and Thinking,” Justice Institute of British
the fingerprint was evaluated.
Columbia, 2017.
• In order to objectively evaluate the process steps [4] A.S. Mbugua. ”Crime Scene Management,” Accessed on:
performed by the each crime scene component Nov. 13, 2019. [Online]. Available: https://www.slideshare.net
/SteveMbugua/crime-scene-management-compiled-52804284.
and their interaction with each other, a number
[5] W.J. Chisum, B.E. Turvey. ”Crime reconstruction,” Academic
of criteria were determined. Press/Elsevier, 2011.
• The problems encountered during the forensic [6] ”90 Criminal Justice Jobs,” 2009. Accessed on: Dec. 5, 2019.
[Online]. Available: https://thebestschools.org /careers/criminal-
fingerprint analysis process were identified and
justice-jobs/.
technology-supported solution suggestions were [7] M.A.M. Mawaheb. ”Principles of Crime Scene Investigation,”
presented to solve these problems. M.S. Thesis, Cairo University, 2019. Accessed on: Oct. 15,
2019. [Online]. Available: https://www.researchgate.net/publica
• To overcome the difficulties, feasible ap-
tion/331745757 Principles of Crime Scene Investigation.
proaches were identified and the applicability [8] R. Julian, S. Kelty, J. Robertson. ”Get it right the first time,
of these approaches were discussed. Critical Issues at the Crime Scene,” Current issues in criminal
justice, 24(1), pp. 25-37, 2012.
• In order to overcome these difficulties, feasible
[9] M. Kaygısız, ”Adli bilimler,” Seckin Press, 2003.
approaches were identified and the applicability [10] D. Çeker. ”Olay Yeri İnceleme ve Çalışmalarında Adli Arkeolog
of these approaches were discussed. ve Adli Antropologların Rolü: Kuzey Kıbrıs ve Türkiye’deki
Güncel Durum,” Antropoloji, (32), pp. 13-21, 2016.
• The disciplines supported by the forensic sci-
[11] B.A. Fisher. ”Techniques of crime scene investigation,” CRC
ences, and the components that compose the press, 2003.
scene were approached from a different point [12] P. Parmar. ”Reconstruction of crime – A review,” International
Archives of Integrated Medicine, 2(10), pp. 49-53, 2015.
of view and the difficulties encountered in eval-
[13] C. Roux, B. Talbot-Wright, J. Robertson, F. Crispino, O. Rib-
uating the process were identified and, aux. ”The end of the (forensic science) world as we know it?
• 14 problems and four challenges were summa- The example of trace evidence,” Philosophical Transactions of
the Royal Society B: Biological Sciences, 370(1674), 2015.
rized.
[14] J.K Roman, S. Reid, J. Reid, A. Chalfin, W. Adams, C.
Knight. ”The DNA field experiment: Cost-effectiveness anal-
Finally, the deficiencies in the current approaches ysis of the use of DNA in the investigation of high-volume
crimes,” 2008. Accessed on: Oct. 19, 2019. [Online]. Available:
used in the CSI processes and the risks that may https://www.ncjrs.gov/pdffiles1/nij/grants/222318.pdf2008.
arise as a result of these deficiencies were empha- [15] ”Forensic Evidence Types, A Simplified Guide To Fingerprint
sized. This study is expected to have a guiding effect Analysis,” Accessed on: Dec. 3, 2019. [Online]. Available:
http://www.forensicsciencesimplified.org/prints/Fingerprints.pdf.
in order to eliminate these risks and to prove the va- [16] F. Gül. ”Ninhidrinin schiff bazı oluşturma özelliğinden faydala-
lidity of innovative approaches to institutions and to narak parmak izi tayini işleminde kullanılması,” Ph.D. disserta-
evaluate the crime scene with an analytical point of tion, Selcuk University, 2014.
[17] Ö. Gültekin. ”Olay yeri incelemesinde karşılaşılan sorunlar ve
view and proposed future directions to researchers çözüm önerileri,” Türkiye Adalet Akademisi Dergisi, 2(2), pp.
and academics for better and faster processing. 473-508, 2011.

110
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

[18] ”Crime scene and physical evidence awareness for non-forensic [35] S.F. Kelty, R. Julian, J. Robertson. ”Professionalism in crime
personnel,” United Nations Office on Drugs and Crime, 2009. scene examination: the seven key attributes of top crime scene
Accessed on: Sep. 3, 2019. [Online]. Available: https://www.uno examiners,” Forensic science policy & management: an inter-
dc.org/documents/scientific/Crime scene awareness Ebook.pdf. national journal, 2(4), pp. 175-186, 2011.
[19] T. F. Kiely. ”Forensic evidence: science and the criminal law,” [36] J. Walvisch. ”Fingerprinting to solve crimes: not as robust
CRC Press, 2005. as you think,” 2017. Accessed on: Oct. 17, 2019. [Online].
[20] I. R. Wagstaff, G. LaPorte. ”The importance of diversity and Available: https:// theconversation.com/fingerprinting-to-solve-
inclusion in the forensic sciences,” National Institute of Justice crimes-not-as-robust-as-you-think-85534.
Journal, 279, pp. 81-91, 2018. [37] ”Evidence Collection Manual,” Accessed on: Dec. 12, 2019.
[21] ”Minimum Requirements For Crime Scene Investigation, [Online]. Available: https://www.crime-scene-investigator.net/
Crime Scene Management,” 2014. Accessed on: Sep. 15, evidencecollectionmanual MT.pdf.
2019. [Online]. Available: http://www.ifsa-forensics.org/wp- [38] A.C. Doyle. ”Methodical Approach to Processing the Crime
content/uploads/2016/09/CSI MRD English.pdf. Scene,” An Introduction to Crime Scene Investigation, pp. 103-
[22] Ş. Parlakyıldız. ”Yapay Sinir Ağları Kullanılarak Parmak İzi 133, 2010.
Tanıma ve Sınıflandırma,” M.S. Thesis, Gazi University, 2014. [39] S. Wilkinson, D. Haagman. ”Good practice guide for computer-
[23] E. Durnal. ”Chemical Visualization of Latent Prints,” based electronic evidence,” Association of Chief Police Officers.
2012. Accessed on: Nov. 23, 2019. [Online]. Available: [40] ”The Basics of Criminal Investigation,” Accessed on: Oct.
https://www.ncjrs.gov/pdffiles1/nij/grants/238008.pdf. 17, 2019. [Online]. Available: http://samples. jbpub.com/978128
4082852/9781284082852 CH04 Dutelle SECURE.pdf.
[24] E.M. Robinson. ”Crime scene photography,” Academic Press,
[41] P. Salicco. ”Staffing a Crime Scene Investigation Unit from
2016.
Sworn to Civilian: A Solution for Law Enforcement Opera-
[25] Y. Martin, M.S. Kumar. ”Forensic Photography: A Review”,
tions,” Journal of Forensic Sciences & Criminal Investigation,
International Journal of Forensic Science & Pathology, 3(9),
4(2), 2017.
pp. 169-171, 2015.
[42] ”The International Association for Identification,” Accessed on:
[26] S. Gouse, S. Karnam, H.C. Girish, S. Murgod. ”Forensic
Nov. 9, 2019. [Online]. Available: https://www.theiai.org/.
photography: Prospect through the lens,” Journal of forensic
[43] M.C. Leone, R. Keel. ”Occupational Stress and Crime Scene
dental sciences, 10(1), pp. 2-4, 2018.
Investigator,” Journal of Law and Criminal Justice, 4(1), pp.
[27] M.T. Miller. ”Crime scene investigation,” Forensic science, pp.
63-74, 2016.
143-164, 2002.
[44] ”ISO/IEC 17020,” Accessed on: Dec. 15, 2019. [Online]. Avail-
[28] Ö.S. Sönmez. ”Bilgisayar destekli parmak izi tanıma sistemi able: https://www.iso.org/standard/52994.html.
tasarımı,” M.S. Thesis, Istanbul University, 2008. [45] ”ISO/IEC 17025,” Accessed on: Dec. 15, 2019. [On-
[29] A.B. Kanbar. ”Fingerprint identification for forensic crime scene line]. Available: https://www.iso.org/ISO-IEC-17025-testing-
investigation,” International Journal of Computer Science and and-calibration-laboratories.html.
Mobile Computing, 5(8), pp. 60-65, 2016. [46] ”Scenes of Crime Examination Best Practice Manual,” ENSFI
[30] M. Ulker, B. Arslan, S. Sagiroglu. ”Evaluation of Fingerprint Scenes of Crime Working Group, Available: http://library
Enhancement Techniques Used by Crime Scene Investigation,” .college.police.uk/docs/appref/ENFSI-BPM-v1 0.pdf.
10th. International Conference On Information Security And [47] M. Walport, C. Craig, E. Surkovic. ”Forensic Science And Be-
Cryptology, pp. 29-37, 2017. yond: Authenticity, Provenance And Assurance,” Government
[31] K.R. Moses, P. Higgins, M. McCabe, S. Prabhakar, S. Swann. Offiice for Science.
”Automated fingerprint identification system (AFIS),” Scientific [48] J.G. Cino. ”Forensic evidence largely not supported by sound
Working Group on Friction Ridge Analysis Study and Technol- science – now what?,” 2016. Accessed on: Dec. 5, 2019. [On-
ogy and National institute of Justice-The fingerprint sourcebook, line]. Available: https://theconversation.com/forensic-evidence-
pp. 1-33, 2011. largely-not-supported-by-sound-science-now-what-67413.
[32] ”Automated Fingerprint Identification System (AFIS) – a short [49] M.N. Singh, S. Joshi. ”Digital image forensics: progress
history,” Accessed on: Dec. 2, 2019. [Online]. Available: and challenges,” 31st National convention of Electronics and
https://www.gemalto.com/govt/biometrics/afis-history. Telecommunication Engineers, 2016.
[33] B.T. Ulery, R.A. Hicklin, J. Buscaglia, M.A. Roberts. ”Accuracy [50] C. Neumann, ”Fingerprints at the crime-scene: Statistically
and reliability of forensic latent fingerprint decisions,” Proceed- certain, or probable?,” Significance, 9(1), pp. 21-25 2012.
ings of the National Academy of Sciences, 108(19), pp. 7733- [51] L. Haber, R.N. Haber. ”Scientific validation of fingerprint
7738, 2011. evidence under Daubert,” Law, Probability and Risk, 7(2), pp.
[34] M.B. Thompson, J.M. Tangen, D.J. McCarthy. ”Expertise in 87-109, 2008.
fingerprint identification,” Journal of forensic sciences, 58(6), [52] S. Bramble, D. Compton, L. Klasén. ”Forensic image analysis,”
pp. 1519-1530, 2013. 13th INTERPOL Forensic Science Symposium, 19, 2001.

111
INTERNATIONAL JOURNAL OF INFORMATION SECURITY SCIENCE
Arslan et al., Vol.8, No.4, pp.88-113

[53] ”Compendium Of Fingerprint Equipments,” 2018. Accessed on: Systems (IEEE World Congress on Computational Intelligence,
Oct. 13, 2019. [Online]. Available: http:// ncrb.gov.in/Bureau pp. 2169-2176, 2008.
Divisions/CFPB/pdf/Miscellaneous/Compendium%20of%20Fin [69] S. Sagiroglu, N. Ozkaya. ”Artificial neural network based
gerPrint%20Equiments-2018.pdf. automatic face parts prediction system from only fingerprints,”
[54] T.F. Wilson. ”Planning for Automated Fingerprint Identifica- Computational Intelligence in Biometrics: Theory, Algorithms,
tion Systems (AFIS) Implementation,” 1988. Accessed on: and Applications, pp. 77-83, 2009.
Nov. 3, 2019. [Online]. Available: https://www.ncjrs.gov/pdf [70] S. Siva, J. Abinaya, S.S. Raja. ”A Cost-Effective Method
files1/Digitization/115419NCJRS.pdf. for Blood Group Detection Using Fingerprints,” International
[55] R. Tredinnick, S. Smith, K. Ponto. ”A cost-benefit analysis of Journal of Advance Study and Research Work, 2(3), 2019.
3D scanning technology for crime scene investigation,” Forensic [71] N.C. Prashantha, K.P. Nabil Abdulla, S. Gurumurthy. ”Age
Science International: Reports, 2019. Determination from Obtained Fingerprint Using 2D Discrete
[56] ”Forenscope Temassız Parmak İzi Tanıma,” Accessed on: Wavelet Transforms and Support Vector Machine,” Interna-
Dec. 12, 2019. [Online]. Available: http://www.forenscope. tional Journal of Advanced Research in Electronics and Com-
com/tr/products/contactless-fingerprint. munication Engineering, 5(5), 2016.
[57] ”The Vampire Tactical Forensic Device,” Accessed on: [72] A.K. Saxena, S. Sharma, V.K. Chaurasiya. ”Neural network
Dec. 12, 2019. [Online]. https: //www.boozallen.com/e/insight based human age-group estimation in curvelet domain,” Pro-
/publication/the-vampire-tactical-forensic-device.html. cedia Computer Science, 54, pp. 781-789, 2015.
[58] ”Kriminal Delil Fotoğraflama Seti,” Accessed on: Dec. 12, [73] R. Bansal, P. Sehgal, P. Bedi. ”Minutiae extraction from fin-
2019. [Online]. https://papilon.com.tr/tr/urunler/kriminal-analiz- gerprint images-a review,” International Journal of Computer
cihazlari/fosko/. Science Issues, 8(5), 2011.
[59] A.K. Jain, Y. Chen, M. Demirkus. ”Pores and ridges: High- [74] M. Aydın, N. Karakuş, O. Çetin, C.M. Solak, D. Bahadir. ”Olay
resolution fingerprint matching using level 3 features,” IEEE Yeri İnceleme Teknikleri Temel Eğitim Kitabı,” Kriminal Polis
transactions on pattern analysis and machine intelligence, 29 Laboratuarları Dairesi Başkanlığı Yayını, 2006.
(1), pp. 15-27, 2007. [75] ”Strengthening Forensic Science in the United States: A Path
[60] J. Li, J. Feng, C.C.J. Kuo. ”Deep convolutional neural network Forward,” 2009. Accessed on: Dec. 12, 2019. [Online]. The
for latent fingerprint enhancement,” Signal Processing: Image National Academic Press, Available: https://www.ncjrs.gov/pdf
Communication, 60, pp. 52-63, 2018. files1/nij/grants/228091.pdf.
[61] ”Big Data architecture delivered successfully,” Accessed on: [76] S.L. Cooper. ”Challenges to fingerprint identification evidence:
Dec. 14, 2019. [Online]. Available: https: //www.hitachivantara Why the courts need a new approach to finality,” Mitchell
.com/en-us/pdfd/case-study/uidai-case-study.pdf. Hamline Law Review, 42, 2016.
[62] R. King. ”World’s largest biometrics database leverages Big [77] G.S. Bumbrah, R.M. Sharma, O.P. Jasuja. ”Emerging latent
Data architecture,” Available: https://www.biometricupdate.com fingerprint technologies: a review,” Research and Reports in
/201512/worlds-largest-biometrics-database-leverages-big-data- Forensic Medical Science, 6, pp. 39-50, 2016.
architecture, 2015.
[63] A.S. Falohun, O.D. Fenwa, F.A. Ajala. ”A Fingerprint-based
Age and Gender Detector System using Fingerprint Pattern
Analysis,” International Journal of Computer Applications,
136(4), pp. 43-48, 2016.
[64] J.A.Y. Hal, D. Kimura. ”Dermatoglyphic asymmetry and sexual
orientation in men,” Behavioral Neuroscience, 108(6), 1994.
[65] M.D Nithin, B.M. Balaraj, B. Manjunatha, S.C. Mestri. ”Study
of fingerprint classification and their gender distribution among
South Indian population,” Journal of Forensic and Legal
Medicine, 16(8), pp. 460-463, 2009.
[66] M.D Nithin, B. Manjunatha, D.S. Preethi, B.M Balaraj. ”Gen-
der differentiation by finger ridge count among South Indian
population,” Journal of forensic and legal medicine, 18(2), pp.
79-81, 2011.
[67] A.M. Badawi, M. Mahfouz, R. Tadross, R. Jantz. ”Fingerprint-
Based Gender Classification,” Image Processing and Computer
Vision, 6, pp. 41-46, 2006.
[68] N. Ozkaya, S. Sagiroglu. ”Intelligent face border generation sys-
tem from fingerprints,” IEEE International Conference on Fuzzy

112