ABSTRACT
Biometrics is a physical, electronic and biological means of providing dependable and accurate
identification of people. It entails data acquisition, storage and subsequent matching at points of
access or identification. Biometrics is an emerging technology that identifies people by their
physical and/or behavioral characteristics and practically requires the person to be identified to
be physically present at the point of identification. In this thesis, efforts to improve biometrics
were made in developing an algorithm that will enhance fingerprint images for effective
matching and rapid identification. The algorithm was based on wavelet processing which
employed Daubechies’ wavelets for decomposition as well as reconstruction. Wavelet smoothing
and denoising were also included in the algorithm.
TABLE OF CONTENTS
TITLE PAGE ii
DEDICATION iii
ACKNOWEDGMENT iv
ABSTRACT v
LIST OF FIGURES vii
LIST OF TABLES viii
TABLE OF CONTENTS ix
CHAPTER ONE: INTRODUCTION 1
1.1 PREAMBLE 1
1.2 DEFINITIONS 2
1.3 BRIEF HISTORY OF BIOMETRICS 5
1.4 BIOMETRIC SYSTEMS 7
1.4.1 IMPORTANCE OF BIOMETRIC IDENTIFICATION 9
1.4.2 APPLICATIONS OF BIOMETRICS 10
CHAPTER TWO: LITERATURE REVIEW 12
2.1 PRELIMINARY 12
2.2 FINGERPRINT IMAGE ENHANCEMENT TECHNIQUES 13
2.3 WAVELET PROCESSING 23
2.4 IMAGE ENHANCEMENT USING WAVELET PROCESSING 32
2.5 FINGERPRINT IMAGE ENHANCEMENT USING WAVELET
PROCESSING 43
2.6 SUMMARY OF WAVELET TRANSFORM 55
CHAPTER THREE: RESEARCH METHODOLOGY 57
3.1 DEVELOPMENT OF FINGERPRINT IMAGE ENHANCEMENT
ALGORITHM USING WAVELET PROCESSING 58
3.2 BACKGROUND INFORMATION ON WAVELET THEORY 58
10
3.3 FINGERPRINT ENHANCEMENT ALGORITHM BASED ON
WAVELETS 63
3.3.1 ACQUISITION OF FINGERPRINT IMAGE 67
3.3.2 NORMALIZATION 67
3.3.3 WAVELET DECOMPOSITION 68
3.3.4 RECONSTRUCTION OF ENHANCED FINGERPRINT IMAGE 68
3.3.5 WAVELET FILTERING 68
CHAPTER FOUR: EXPERIMENTS 69
4.1 DISCUSSION OF EXPERIMENTS AND PROCEDURE 69
4.2 QUANTIFICATION OF THE DEGREE OF ENHANCEMENT 70
4.3 ANALYSIS OF RESULTS OF FINGERPRINT IMAGE ENHANCEMENT
ALGORITHM BASED ON WAVELET PROCESSING 73
4.4 COMMENTS ON IMPROVEMENTS BY WAVELET PROCESSING 74
4.4.1 SPECIFIC APPLICATION AREAS FOR DEPLOYMENT
OF THE ALGORITHM 74
4.5 LIMITATIONS OF BIOMETRIC SYSTEMS 75
CHAPTER FIVE: 77
5.1 RECOMMENDATIONS 77
5.2 CONCLUSION 78
References: 79
Appendix A-
CHAPTER ONE
INTRODUCTION
1.1 PREAMBLE
Today’s world has been greatly modified by technology in the efforts of man to deal with
his environment in order to tame it to suit his needs; with particular reference to
electronics. Virtually, all aspects of our modern civilization depend on electronic
[including hardware and software] systems for smooth and effective operations of daily
activities. Ranging from military, communications, government, education, banking,
automobiles and manufacturing to the space industry; all rely on electronic systems for
various applications like process automation, control and monitoring systems, security,
automated parking control, 24 hours surveillance systems, traffic collision avoidance
system (TCAS) of aircrafts and on board data handling subsystems of satellites (OBDH).
With the dynamic growth of the world’s population, global economic meltdown, food
scarcity, and unemployment; all these factors have led to significant increase in violent
crimes globally making security a great challenge to all and sundry.
Governments, business owners, the organized private sector and educational institutions
all require a robust, rugged and dependable means of recognizing and verifying peoples’
identities at various points of access to specific areas physically and electronically.
In order to ensure that intruders, impersonators are promptly fished out prior to breaching
the security of these valued resources, information, assets and vaults; a robust and
effective means of identifying and verifying peoples’ identities is paramount. The search
for such systems culminated recently in biometrics.
In the late nineteenth century, it was discovered by Johannes Evangelista Purkinje a
Czechoslovakian physiologist [1] that no two human beings share the same fingerprint
even Siamese twins! This was a revolutionary breakthrough in the world of law
enforcement as investigators now had a valid and reliable means of hunting down as well
as precisely identify criminals [1].
12
Biometrics is an emerging technology that is used to identify people by their physical
and/or behavioural characteristics and practically requires the person to be identified to
be physically present at the point of identification [2].
Biometric identification relies on those characteristic features of humans that are
inherently unique to individuals such as fingerprints, palmprints, retina, iris, hand
geometry, facial features, and ears. Another category of biometrics utilize behavioural
characteristics like speech [or voiceprint], lip movement, keystroke dynamics, gait,
gesture, signature [2] and thermal emissions. These features are so distinct that no two
humans on earth even Siamese twins possess exactly the same replicas! A physiological
or behavioural characteristic that is unique to an individual is referred to as biometric
measurement [e.g. voiceprints and fingerprints] which has the capability to reliably
distinguish between an authorized person and an imposter [3].
Biometric systems capture these features as digital data using various devices like digital
cameras and scanners, store them on a database where they serve as templates, retrieve
the templates as quickly as possible when required for verification or authentication at
various points of access. For a fingerprint based biometric system, a typical verification
phase involves two stages: enrolment and verification.
Enrolment: this stage capture’s the user’s fingerprint; its distinctive features [minutiae]
are extracted and stored as a template.
Verification: during the verification stage, a new fingerprint is acquired and compared to
the stored template to verify the user’s claimed identity [4].
1.2 DEFINITIONS
Fingerprint: A fingerprint is the reproduction of a fingertip epidermis, which is
generated when a finger is pressed against a flat surface or any other surface. The main
structural characteristic of a fingerprint is the pattern of interleaved ridges or ridgelines
and valleys which often run in parallel. [5] Defined fingerprint as a unique pattern of
ridges and valleys on the surface of fingers of an individual.
Fingerprint patterns exhibit one or more regions where the ridgelines assume particular
shapes. Figure 1 shows a fingerprint image taken from FVC 2002 indicating minutiae,
13
core and delta points. These regions are called singularities or singular regions and they
form the bases upon which fingerprints are classified into three different categories:
i. Loop
ii. Delta
iii. Whorl
Figure 1: A whorl type fingerprint showing core points, Delta point and minutiae- source [6]
The singular regions are generally denoted by п, Δ and О shapes respectively. Singular
regions are commonly used for fingerprint classification, which is assigning a given
fingerprint to a class among a set of distinct classes with the sole aim of simplifying
search and retrieval.
Algorithm: Algorithm is defined as a step by step approach of executing a specific task.
For any given task, several algorithms may exist but the results of such algorithms must
be the same.
Algorithm can be thought of as a limited set of precise instructions which when executed
in a predefined manner attains its objective. ‘An algorithm is an effective method
expressed; as a finite list of well defined instructions for calculating a function’ [7].
In computer science, an algorithm is defined as a logical, systematic, computational
approach that is adopted to solve programming tasks. It is a step by step process that
software engineers apply in the design of series of instructions a computer must execute
to produce the desired output.
Minutiae
Delta
Core points
14
Enhancement: It is a process that improves the quality, attribute, features or
effectiveness of any object. To enhance simply means to add value in order to increase
perception.
Centre Point Core: Is defined as the position of the north-most loop singularity or as the
point of maximum ridgeline curvature for fingerprints belonging to the arch class [see
figure 1].
Orientation Image: Is a discrete matrix whose elements denote local orientation of the
ridgeline pattern of a fingerprint.
Local Ridgeline Frequency: Is defined as the number of ridges per unit length.
Minutiae: Are ridgeline discontinuities, local characteristics of the pattern that are stable
and robust to fingerprint impression conditions [8]. Minutiae are distinctive features
adopted by most fingerprint based biometric systems for identification and verification;
they may be classified into several types such as:
i. Termination- the point where a ridgeline suddenly ends.
ii. Bifurcation- the point where a ridgeline divides into two.
iii. Island
iv. Dot
v. Lake
A minutia point may be defined by its type, the x- and y- coordinates and the direction θ.
The ISO/IEC 19794-2:2005 standard specifies the data formats for minutiae based
fingerprint representation. The fingerprint minutiae record format defines the
fundamental data elements used for minutiae based representation of a fingerprint and
optional extended data formats for including additional data such as ridge counts and
singularities location.
Ridge: ‘A ridge is defined as a single curved segment’ of a fingerprint’s pattern [2].
Multimodal Systems: As the name indicates, they are biometric systems that utilize two
or more biometric characteristics for recognition. Multimodal systems generally require
integration schemes as they employ characteristics for identification and verification of
claimed identity.
15
Unimodal Systems: These are biometric systems that are based on a single biometric
feature such as fingerprint or facial feature for recognition. They are generally less
accurate than multimodal systems and also more susceptible violation.
1.3 BRIEF HISTORY OF BIOMETRICS
Research efforts made by various scholars with sponsors from different stakeholders to
provide reliable answers to questions such as these:
i. Does Mr. Obi have administrator privileges on this network?
ii. Is Chima authorized to access this facility?
iii. Has Musa been convicted of a crime prior to this interview?
iv. Is Iniobong who he says he is?
Culminated in Biometrics
Traditionally, organizations and governments have answered questions such as these by
relying on the information the individuals possessed such as passwords or what they had
like identity cards. These, have major setbacks as they can be easily stolen, manipulated
or forged! These limitations provided an opportunity for improvements as well as total
replacements; which led to researches on robust automated systems that can effectively
identify people, verify identities as well as pose a great barrier to any one that attempts to
falsify his/her identity. A need for a system that will make personal recognition based on
who people were was paramount.
Humans have for generations identified each other with various body characteristics like
voice, face, stance, mannerisms and gaits. These characteristics are particularly unique to
each individual; hence any means of identification that can incorporate any of these
features is likely to be foolproof and failsafe.
In the mid 19th century, Alphonse Bertillion, chief of the criminal identification division
of the police department in Paris, developed and practiced the idea of using various body
measurements (for example: height, length of arms, feet, and fingers) to identify
criminals [9].
It was discovered in the late 19th century that no two human beings had the same fingers;
just as Alphonse Bertillon’s practice of body parts measurement was gaining popularity.
16
This revolutionary breakthrough discovery became the prince of law enforcement as
investigators now had a definitive means of identification at their disposal [1].
After this discovery, many major law enforcement departments adopted the concept of
‘booking’ criminals’ fingerprints and storing them in databases or card-files. As the
technology grew, police acquired the ability to ‘lift’ leftover, typically fragmentary
fingerprints from crime scenes (commonly called latents) and match them with
fingerprints in the database to determine criminals’ identities [9].
The English word biometric originated from two Greek words bios and metrikos that
mean life and measure respectively. Biometrics is an emerging technology that
automatically recognizes people based on their physiological and/or behavioural
characteristics.
In 1892, Francesca Rojas an Argentine woman made history as the first killer to be
caught using fingerprint evidence [1]. A fingerprint is potentially worthless unless it can
be linked either to a suspect or a database entry. This implied long hours of manually
searching card indexes in order to find a match for the crime scene fingerprint;
fortunately, computerization techniques [image processing algorithms and DSP
techniques] have changed this practice for good. An example of such algorithm is
indexing fingerprints using minutiae as presented in [10]. The integrated automated
fingerprint identification system (IAFIS) is software that contains a huge and expanding
database of fingerprint records, and manual indexing which may have taken weeks or
even months to match a fingerprint is now accomplished within seconds or minutes
depending on the size of the database.
Each fingerprint is unique; they develop on a foetus at about six months and remain
unchanged until the body decomposes after death! Johannes Evangelista Purkinje a
Czechoslovakian physiologist (1787-1869) was the first to recognize and describe the
nine basic patterns of fingerprints [1].
The earliest recorded practical application of fingerprints as a means of identification was
in 1858, when a young administrative officer in India William Herschel, adopted the
local Bengali practice whereby illiterate workers ‘signed’ for their pay by leaving a
thumbprint [1]. This was done to stamp out pension frauds.
17
In 1879, a Scottish physician working in Japan, Henry Faulds became involved in what is
generally considered to be the first crime solved through the use of fingerprints; when a
burglar making his escape from a Tokyo house left a dirty handprint on a white washed
wall. When the police quickly arrested a suspect, Faulds inspected the man’s handprint
and declared him wholly innocent; much to the amusement of sceptical investigators.
Three days later Faulds was vindicated, as another felon admitted to the burglary and his
handprint also matched exactly the one left on the wall [1].
Faulds published his findings in Nature, the British scientific journal in 1880, sparking a
long running feud with Herschel; who still regarded himself as the father of
fingerprinting [1].
In 1892, Sir Francis Galton another English scientist, turned theory into practical reality
when he laid out the ground rules for a basic classification system for fingerprints. He
separated the most commonly observed feature into three categories:
i. Arches
ii. Loops
iii. Whorls
His iconic work, Finger Prints (1892), lit the beacon but it was another Indian police
officer Edward Henry that completed the task of fingerprint classification.
Biometrics initially came into wide usage for law enforcement and legal purposes such as
identification of criminals and illegal aliens, security clearances for employees in
sensitive jobs, paternity determinations, forensics, positive identification of convicts, and
prisoners. However, in recent times, many civil and private establishments apply
biometrics to establish personal recognition and prevent identity thefts.
In 1956, Alphonse Bertillon conceptualized the idea and commercially implemented the
procedure of applying biometrics [body measurements] in crime investigation and
prosecution.
1.4 BIOMETRIC SYSTEMS
These are systems that identify, verify, authenticate and grant access to restricted
resources or facilities based on distinctive human characteristics such as voiceprints,
fingerprints, facial features and iris recognition. Generally, biometric systems are of two
18
categories viz: unimodal [are biometric systems that are based on single biometric
characteristic such as fingerprint recognition systems] and multimodal [refer to biometric
systems that rely on multiple biometric features e.g. voiceprint & iris recognition
systems] biometric systems.
A biometric system is primarily a pattern recognition system that recognizes an
individual based on a feature vector obtained from a specific and unique physiological or
behavioural characteristics that the individual possesses [1].
Figure 2 shows a generic architecture of a biometric system. It can be seen from the
diagram that a biometric system can function in either of two modes namely:
i. Verification mode
ii. Identification mode
Verification mode: In this mode, the biometric system authenticates a user’s claimed
identity by matching the captured biometric characteristic against his/her biometric
template which has been previously enrolled. In this mode, a biometric system is required
to answer the question, is Mr. A truly Mr. A. It provides identity verification using
positive recognition in which the system performs a one to one comparison to determine
whether the claimed identity is true or false. An example of this type of biometric system
is electronic employee attendance register.
Identification mode: Recognition of an individual in identification mode is performed
through a one to many searches. The biometric system searches its entire database of
templates for a match of the individual’s biometric information. If the individual’s
biometric data has not been enrolled previously, the system fails to identify him/her. The
question that is answered in this mode is who is Mr. B? This is a form of negative
recognition whereby the system establishes if Mr. B is who he implicitly is or explicitly
denies being. An example of a biometric system that functions in negative recognition
mode is fingerprint based access door control. The essence of negative recognition is to
prevent a single individual from using multiple identities as evident in Nigeria’s 2011
voter registration exercise. It is also possible to apply identification in positive
recognition for convenience when the user is not required to make an identity claim [9].
19
1.4.1 IMPORTANCE OF BIOMETRIC IDENTIFICATION
Conventional methods applied in personal identification like tokens, pins, keys and
passwords provide only positive recognition whereas biometrics can provide both
negative and positive recognition. The benefits of biometric identification cannot be
overemphasized, let us highlight just a handful as an insight into the immense features
that characterise biometrics:
i. Rapid Identification: Biometric systems provide rapid as well as reliable
identification of individuals at various points of access that are sensitive and
require restricted access.
ii. Employee Attendance Register: Biometric based staff attendance register
provides a robust and dependable means of taking staffs’ attendance and does not
give room for manipulation. It can also serve as a reference for
Capture Signal
processing
Decision
Data storage Matching
Matching
? ?
Presentation Threshold
Sensor
Template creation
Quality control
Feature extraction
Re-acquire
yes/no list
Enrolment
Verification
Identification
Figure 2: Generic architecture of a biometric system: source [11].
Templates
20
awards/punishments of staff based on punctuality and dedication to duty as a
sequential order of attendance is logged and stored in a database.
iii. Verification: Biometric systems provide reliable means of verifying peoples’
identities of particular reference is at points of bank transactions; where account
holders will be required to be distinguished from third parties.
iv. Forensics: It has become imperative to preserve the integrity of crime scenes in
order to accurately acquire evidence(s) that assist in tracking down and convicting
criminals. Such evidences when garnered will require a robust, reliable and
accurate system for analysis to provide failsafe identification. Biometric systems
provide such platforms.
v. Archaeological Discoveries: Biometric systems are applied in the analysis of
prehistoric manuscripts for handwriting recognition and also traceability with
respect to the era and cultures in which such documents were written.
vi. Security: Biometric systems are generally less susceptible to hackers as is the
case with traditional identification technologies. Biometric data are unique and
since they require the physical presence of the individual to be identified, security
risks are minimized. Biometrics cannot be easily shared, duplicated, lost, or stolen
as is the case with the conventional technologies.
vii. Convenience: They are very convenient to use as they cannot be forgotten or lost,
forged, repudiated without compromising their accuracy or integrity as they
ensure that the user is always present at the point and time of recognition.
1.4.2 APPLICATIONS OF BIOMETRICS
Biometrics can be applied broadly in four major categories:
i. Government applications: This category of application lists various
functions/responsibilities a government should offer its citizens; for instance:
issuance of national identity cards, voters’ registration, issuance of drivers’
licenses, prison services, border monitoring and control, passport control and
immigration services as well as administering aid services.
ii. Military applications: Areas that biometrics can be deployed in the military
include: personnel identification, clearance level or rank regulated resources (such
21
as access to Amory, clandestine mission authorization, command and control
centres for active operations), hardware control and authorization, missile launch
and facility control.
iii. Commercial and civil applications: Numerous areas of biometrics’ application
abound commercially, examples include: computer access control, electronic
attendance systems, computer network access control, automated teller machines,
e-commerce, internet access, physical access control, mobile phones, personal
digital assistants, electronic data management systems, debit and credit cards
authentication, medical and official records management and distance learning
and so on.
iv. Forensic applications: This category generally deals with medical and scientific
methods of crime solving; examples include identification of corpse, criminals’
and terrorists’ identification, missing children, crime scene investigation.
Figure 3: Examples of biometric systems application (a) Digital Persona’s fingerprint verification
system provides personal recognition for computer and network login. (b) Indivos manufactures a
fingerprintbased point-of-sale (POS) terminal that verifies customers before charging their credit
cards. (c) BioThentica’s fingerprint-based door lock restricts access to premises. (d) The Inspass
immigration system, developed by Recognition Systems and installed at major airports in the US,
uses hand geometry verification technology: source-[9].
22
Do you need help? Talk to us right now: (+234) 08060082010, 08107932631 (Call/WhatsApp). Email: [email protected].
IF YOU CAN'T FIND YOUR TOPIC, CLICK HERE TO HIRE A WRITER»
