UNIT 4 BIOMETRICS* Biometrics

Contents

4.0 Introduction
4.1 Biometrics: Definition and System
4.1.1 Advantages
4.1.2 Disadvantages
4.1.3 Applications of Biometrics
4.2 Types of Biometrics
4.2.1 Finger Print Biometrics
4.2.2 Iris Patterns Biometrics
4.2.3 Retinal Pattern Biometrics
4.2.4 Facial Recognition Biometrics
4.2.5 DNA Biometrics
4.2.6 Handwriting and Voice Biometrics
4.3 Summary
4.4 References
4.5 Answers/Hints to Check Your Progress

LEARNING OBJECTIVES

After the completion of this unit you will be able to understand:

 biometrics, its scope and role in forensic anthropology;

 the use of human characteristics in personal identification with the help

of computer technology; and

 characteristics and the methodology used in biometrics for the individual

identification.

4.0 INTRODUCTION
‘Forensics’ means ‘for the court’ and it relies on the evidence and
identification. You are well aware that forensic analysis is dependent on the
evidence collected from the crime scene for the identification of the criminal.
With the increase in crimes new techniques for investigation like Biometrics
based on artificial intelligence are being developed.

Biometrics consists of two words ‘Bio’ meaning ‘life’ and ‘metric’ is the
measurement or quantitative based recognition and identification. In
simple words if you have to choose a T-shirt for your blue jeans what you

*
Contributor: Dr Kumud Sarin, Program Director at Bioinformatics Institute of India,
Noida. 73
Human would do? You would choose from a pile and match, much the same way
Identification:
Establishing in Biometrics where we compare given pattern from the existing ones in
Identity-I the data base. Basically, it is related to brain recognition with its stored
observations and is faster with more accuracy.

4.1 BIOMETRICS: DEFINITION AND SYSTEM

Biometrics, in simple terms is identification of humans by their traits related
to different aspects of physiology, chemistry or behaviour. Biometrics is
defined as the applicable means of identifying and authenticating individuals
in a reliable and fast way through the use of unique biological characteristics
and technologies that lead to the acceptable way of solving the crime.
Biometrics includes two types of characteristics:

 Physiological: face, fingerprint, hand geometry and iris recognition, and

 Behavioural: gait, odour, signature and voice.

Figure1: Biometric Features

Source: http://www.biometrika.it/eng/wp_biointro.html

Biometric system use the following human characteristics as identifiers:

Common:

Fingerprint Recognition
Face Recognition
Speaker Recognition
Iris Recognition
Hand Geometry
Signature Verification

Others:

DNA
Retina Recognition
Thermo Grams
Ear Recognition
74
 Skin Reflection Biometrics

Lip Motion
Body Odour
Brain Wave Pattern
Footprint and Foot Dynamics

Biometric system consists of device which works as a system shown in figure 2.

Figure 2: Biometric System

Source: https://en.wikipedia.org/wiki/Biometrics

From diagram you can see that a sensor recognises the information, then
during pre-processing, it extracts information with template for matcher and
identifies the information.

4.1.1 Advantages
Biometrics with any device, either system or characteristics are used with a
common aim to collect the information and compare for identification. The
advantages are:
 Universally found in all individuals and unique to differentiate one
individual from another,
 Permanent, allowing for change over time
 Recordable (with or without consent)
 Measurable for future comparison
 Forgery-proof (a face, a fingerprint)
 Accuracy, improved security systems and quick authentication

75
Human 4.1.2 Disadvantages
Identification:
Establishing
Identity-I Biometric system has advantages nevertheless also has disadvantages like
any other system. Some of the disadvantages are:

• Physical disabilities caused by injury, etc. leads to change the system

• Expensive
• Physical traits can’t be altered
• Scanning is not reliable e.g. in retina scanning
• Software issues like power issues
• Remote access is lacking and security breaches

4.1.3 Applications of Biometrics

You would like to ask the question that when Biometrics has disadvantages
then why it is developing. Biometrics has lot of commercial and daily life
applications which creates innovative new systems which are precise and less
erroneous. It is used in the:

Identification and Authentication

Legal applications: Justice and law enforcement
Government applications
o Border control and airport
o Healthcare and Medical
Commercial applications
o Security
o Attendance and Time Management
Finance applications of biometric technology, financial identification,
verification, and authentication in commerce to make banking,
purchasing, and account management safer and more convenient and
responsible.
Mobile
Eye movements tracking applications
Automotive industry
Screen navigation
Aviation
Detection and recognition of dynamic shapes
Forensic Science Applications
Physical and logical access (owner/user/employee/contractor/partner
identification) identification
Commercial applications (consumer/customer ion)
76
Check Your Progress 1 Biometrics

1) Define biometrics and state its importance in forensic science.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

4.2 TYPES OF BIOMETRICS

Biometrics

Behavioural Biometrics Physiological Biometrics

Behavioural Biometrics is the scientific study of body function and is

divided as:

Signature Recognition: The biometric signature recognition system

verifies an authorised user for authentication and authorisation.

Voice Recognition: Voice is a behavioural as well as a physiological

trait that depends on the physical structure of the throat and mouth and
helps to distinguish the speaker as per the speech.

Keystroke: Identification on the basis of striking the keyboard and its

rhythm.

Physiological biometrics is based on physical characteristics like, ears, eyes,

iris, fingerprints, and is divided as:
Ear authentication: It is based on sound waves to determine the way of
the ear canal of an individual.
Eye vein recognition: Biometric identification method by employing
pattern formed by veins on the sclera (the white outer coating of the eye).
Facial recognition: 3D structure of face recognition.
Finger vein recognition: Blood vessels pattern recognition under the
skin of fingers.
Fingerprint recognition: Fingerprint recognition is comparing the
finger prints with available data as no two persons can have similar
fingerprints.
77
Human DNA matching: The identification based on the analysis of segments
segm
Identification:
Establishing from DNA.
Identity-I
Footprint and foot dynamics: Footprintootprint is a unique physiological
character with distinctive properties useful in personal identification

Gait recognition: A gait is defined as the cyclic and coordinated

combination of movements for human loc locomotion
omotion and a person can be
identified by observing his/her gait as a distinguishable feature.

4.2.1 Finger Print Biometrics

4.2.1.1 Introduction
Fingerprint is defined as featured pattern of one finger of ridges and valleys
on the surface of a fingertip. The uniqueness in the fingerprint of every
individual is based on local ridge characteristics not evenly distributed.

Figure 3: Finger Print

Source
Source: https://www.dreamstime.com/stock-images-fingerprint
fingerprint-image8262454

The differentiation of the fingerprints is minutiae i.e. prominent local ridge

characteristics or abnormal points on ridges as mentioned below:
1) Ridge ending: the point where a ridge ends abruptly.
2) Ridge bifurcation: the point where a ridge forks or diverges into branch
ridges.

Figure 4: Ridge and Valley

Source:
Source:https://www.semanticscholar.org/paper/Fast-Enhancement
Enhancement-Algorithm-with-
Improved-Clarity
78
 Biometrics

Figure 5: Three Levels of Finger Print Classification

Source:https://www.researchgate.net/publication/281372646_A_Minutiae_Count_Based_M
ethod_for_Fake_Fingerprint_Detection/figures?lo=1

The uniqueness of fingerprints is due to furrows and friction ridges present

on fingers or thumb. On the basis of ridge pattern finger prints are classified
into three types:
Whorl: The spiral patterns giving an appearance of whirlpools is called
whorl and has four types- double loop i.e. S –shaped pattern, concentric
plain circles, loop along with whorl i.e. central pocket loop and irregular
shaped whorl.
Loop: The patterns that recurve back to form radial or ulnar loops
dependings on the pointing direction i.e. towards ulna or radius.
Arch: A wavy pattern i.e. tented ( arch rise at sharp point) or plain arches.

Figure 6: Pores and Ridges

Source:https://www.semanticscholar.org/paper/Pores-and-Ridges%3A-Fingerprint-Matching
79
Human 4.2.1.2 Finger Print Biometrics Definition
Identification:
Establishing
Identity-I Finger Print Biometrics is defined as the method of identifying finger print by
employing the software on the basis of pattern of finger and location of the
minutiae.

Figure 7: Six Patterns

atterns of Fingerprint
Source: https://www.popsci.com/science/fingerprint-patterns-genetics
https://www.popsci.com/science/fingerprint

Structure of Minutiae
Minutiae:
Crossover two ridges cross each other
Core center
Bifurcation ridge separates
Ridge ending end point
Island small ridge between two spaces
Delta space between ridges
Pore human pore

Figure 8: Structure of Minutiae

Source: https://www.researchgate.net/
80
4.2.1.3 Fingerprint Recognition (Physiological Trait) Biometrics

The process of comparison between the known fingerprint against another in

order to determine that it is of the same finger or palm is known as fingerprint
recognition and consists of two parts.

Fingerprint Verification: to verify the authenticity of one person by his/her

fingerprint.

Fingerprint Identification: to specify one person’s identity by his/her

fingerprint(s).

Figure 9: Fingerprint Recognition

Source: https://electricalfundablog.com/fingerprint-recognition-system-architecture-works

Three Stages of Fingerprint Recognition System

1) Fingerprint Acquiring Device: Optical or semiconductor sensors are

employed for acquisition with high efficiency and acceptable accuracy
while exceptions are dry or dirty fingers.

2) Minutia extractions:

Uses the ridge endings and bifurcations on a person’s finger to plot points are
known as Minutiae. The variations in the number and location of minutiae
occurs from person to person, finger to finger in any particular individual

Finger Finger+ Minutiae Minutiae

Figure 10: Extraction of Minutiae
Source: https://biometrics.mainguet.org/types/fingerprint/fingerprint_algo.htm

81
Human This level has three stages pre-processing, minutiae extraction and post
Identification:
Establishing processing stage.
Identity-I

Verify
Capture Extraction Comparison
individual?

No Yes
Access denied Access to
cannot application sign
sign record records

Scan left Thin image to a Sample minutia

index finger single pixel graph

Acceptable ?
score
Ending
minutiae

Bifurcation
Identify minutiae
minutiae

Reference minutia
graph for individual

Minutia
graph

Figure 11: Finger Print Scanning

Source: https://slideplayer.com/slide/13213653/

3) Minutia Matching
Automatic finger print verification is achieved with minutia matching (point
pattern matching) instead of a pixel-wise matching or a ridge pattern
matching of fingerprint images. First of all, minutia matcher selects two
minutiae as reference minutia pair after that matching of the ridges is done
and finally on the basis of the ridge matching the image are aligned and
82 matching of all minutiae is overseen.
 Biometrics

Figure 12: Finger Print Attendance and Minutiae Matching

Source: https://slideplayer.com/slide/13213653/

4.2.1.4 Advantages and Disadvantages of Fingerprint Biometrics

Advantages

 Fingerprint pattern remain stable throughout the lifetime, unique in nature.

 It is easily analysed and can be compared.
 Inexpensive device and oldest form of biometrics.

Disadvantages

• Wet or moist fingers, cut fingers, or dirt or grease can sometimes affect
the authentication process.
• It is not right tool for those persons who work in chemical labs.

Applications of Finger Printing

Banking Security - ATM security, card transaction

Physical Access Control (e.g. Airport)
Information System Security
National ID Systems, Passport control (INSPASS), Voting
Prisoner, prison visitors, inmate control
Identification of Criminals and missing children
Secure E-Commerce (Still under research)
A person’s biometric identity is on the basis of one–to-one match or on other
security measures like password, PIN etc.

Figure 13: Fingerprint Biometrics

Source: https://slideplayer.com/slide/13213653/
83
Human Check Your Progress 2
Identification:
Establishing
Identity-I 2) Explain fingerprint recognition with applications and advantages.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

3) Draw different patterns of fingerprint.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

4.2.2 Iris Patterns Biometrics

Eyes are important part of the individual and iris or the coloured part of the
eye which has thick and thread-like muscles for Biometric recognition as
shown in the figure 14.

Figure 14: Iris Structure

Source: https://en.wikipedia.org/wiki/Iris_(anatomy)
84
 Biometrics
Iris is made up of pigmented elastic tissue with an adjustable circular
opening in the centre. The muscular pattern and colour is unique in every
individual even the right and left eye is different thus it becomes a reliable
biometric recognition character. Therefore you have to remember that the
colour and muscular pattern is a feature for Iris Recognition along with
distant technology applicable.

Iris recognition is an automated method of biometric identification by

employing mathematical pattern recognition techniques on the images of the
iris of an individual's eye or unique complex random patterns.

Figure 15: Eye Structure

Source: https://www.merckmanuals.com/home/eye-disorders/biology-of-the-eyes/structure-
and-function-of-the-eyes

4.2.2.2 Iris as Biometric

Further you have to understand that the function of muscles is to provide
shape to pupil and light that enters the eye. The uniqueness of the folds of
muscles results into biometric authentication to confirm the identity with
incredible accuracy. Other essentials you have to note is liveness detection
like blink for scan enhance its accuracy and security. The technology has
advantage of its applicability from a distance also. Note that the iris pattern is
not subject to alteration with age, eye redness or alcohol consumption. Iris
recognition can be used as Biometric because iris is the coloured portion of
the eye surrounded by the pupil and its pattern is the result of meshwork of
muscle ligament while colour and contrast are due to pigmentation.

4.2.2.3 Iris Recognition (Physiological Trait)

The recognition of Iris is carried out on the basis of the following traits:

Iris scanning measures the iris pattern in the coloured part of the eye.
Analysis of the iris of the eye, which is the coloured ring of tissue that
surrounds the pupil of the eye.
85
Human Based on visible features.
Identification:
Establishing Widely regarded as the most safe, accurate biometrics technology
Identity-I
High speed, high accuracy.

Figure 16: Eye

Source: https://www.merckmanuals.com/home/eye-disorders/biology-of-the-eyes/structure-
and-function-of-the-eyes)

4.2.2.4 Iris Recognition Systems

Biometric Iris recognition is based upon the working of camera with the
mathematical analysis of the random patterns. Imagine Iris as your phone
camera where camera and light combine to take a photo. In the same way iris
camera carries out a person’s identity by combining computer vision, pattern
recognition, statistical inference and optics.

Figure 17: Eye Structure from Front

Source:https://www.dreamstime.com/structure-eye-anatomical-external-structure-iris-
cornea-eye-pupil-structure-eye-anatomical-image168611026

86
This is how it happens: Biometrics

1) A person stands in front of the iris identification system with a distance

of one or three feet so that the wide angle camera calculates the position
of their eye.
2) The function of the second camera is to zoom on the eye and takes a
black and white image.
3) The focus of the iris overlays a circular grid on the image of the iris thus
identifying the light and dark areas, like an “eye print”.
4) Fake eye prevention or fool system is done by the devices which vary the
light shine into the eye and watch for pupil dilation.

Image Iris
Operator Image
Acquisition Localization

Bar Code Stored in Numeric Digitizing

Pattern
Computer Code Image
Matching
Database

Figure 18: Block Diagram of Iris Recognition

Source:https://www.researchgate.net/publication/224597357_Iris_Segmentation_
Using_Geodesic_Active_Contours/

4.2.2.5 Iris Recognition Function Steps

Four functional steps for Iris recognition are

1) Image Acquisition
2) Pre-processing
3) Image Analysis
4) Image Recognition

1) Image Acquisition

This is an important step as image acquired would lead to recognition and the
characteristics required are:

 Good contrast,
 high quality image and
 high illumination.

Camera should be able to capture volume width on the order of 20–30 cm

and more than 200 pixels across the iris for good quality as 100–150 pixels 87
Human results for marginal quality. Daughman’s (1993) Algorithm described the
Identification:
Establishing methodology for image analysis i.e. Iris Segmentation as described: Iris is
Identity-I mapped into a rectangle in normalised polar coordinate system and
segmentation normalises for scale change and pupil dilation.

Figure 19: Mapping of Iris

Source:https://www.researchgate.net/publication/224597357_Iris_Segmentation_Using_Geo
desic_Active_Contours

Segmentation: Boundary Detection

Segmentation is simplified by modelling the inner and outer iris

boundaries as non-concentric circles.

For each boundary, we must find 3 parameters: x and y of centre, and

radius r.

Search Criteria is based on intensities along with an expanding circular

contour that become suddenly brighter (from red circles to green circles).

First Step: Iris Segmentation

88
“Unwrapping” the iris Biometrics

Outer boundary (with sclera)

Inner boundary (with pupil)

Figure 20: Boundary Detection Example
Source:https://casvisualfoundation.wordpress.com/2015/03/24/tuesday-excursion-research-
and-experiment-4/

2) Pre-processing

Next step is to classify whether the image acquired is useful or not which
means the irrelevant parts e.g. eyelids or pupil and the pre-processing would
remove the effect of the spots /holes lying on the pupillary area. This process
also transforms the true colour into intensity image. The two steps of pre-
processing are:

a) Iris Localisation

The outer and inner boundary of the Iris is considered as two circles
that are not co-centric thus inner boundary between the pupil and the
iris is detected. To detect the outer boundary due to its low contrast
between the two sides is found by maximizing changes of the
perimeter- normalised along the circle.

b) Edge Detection

The usage of Edge Detection is to evaluate complex object

boundaries by marking potential edge point corresponding to places
in an image where rapid change in brightness occurs and edge is
defined by the discontinuity in grey values thus an edge separates
two distinct objects.

3) Image Analyses

Image analysis visualises the features of Iris and digitised into a 512 byte
(4096 bits) Iris Code record. Half Iris Code describes the features and another
half controls the comparison process.

4) Image Recognition

The database is built by storing the different Iris Code. You can apprehend
that the recognition is same as going to the market to buy a T-shirt of
89
Human particular colour by searching and comparing. Same way an iris is presented
Identification:
Establishing at a recognition point and process is repeated and compares it to every file in
Identity-I the Database Recording of Identities thus the Iris Code record is not stored.

Image Acquired from distant Iris is detected from Image

video/image/database

Iris pattern coded into Binary

Matched with other codes and Code
decision taken

Figure 21: The Process Overview

4.2.2.6 Advantages of Iris Technology

Advantages of the technology are mentioned below:

 Accurency: Uniqueness of iris patterns hence improved accuracy.

 Highly protected: It is the internal organ of the eye.
 Stability: Persistence of iris patterns.
 Non-invasive: Relatively easy to be acquired.
 Unique: The probability of two rises producing the same code is nearly
impossible.
 Flexible: Iris iris recognition technology easily integrates into existing
security systems or operates as a standalone.
 Stability: Patterns apparently stable throughout life.
 Reliable: A distinctive iris pattern is not susceptible to theft, loss or
compromise.

4.2.2.7 Disadvantages of Iris Technology

Iris Biometrics disadvantages are listed below:

Small target (1 cm) to acquire from a distance (1m).

Alcohol consumption causes deformation in Iris pattern.
Illumination should not be visible or bright.
Obscured by eyelashes, lenses, reflections.
Difficult to capture an image of handicap people sitting on wheel chair as
cameras are usually attached on the wall therefore can capture an image
up to a certain height.
Costlier iris recognition systems.
Authentication is affected with glasses or direct sunlight.
90
4.2.2.8 Applications of Iris Recognition System Biometrics

Iris recognition is playing an important role:

in providing unique identification to the individuals for the bank

accounts aadhar card etc.
as living password for computer login.
as living passport for national borders security.
driving licenses and personal certificates.
internet security, control of access to privileged information.
premises access control (home, office, laboratory).
anti-terrorism (e.g. security screening at airports).
financial transactions (electronic commerce and banking).
iris password for the bank cash machine accounts.
credit-card authentication.
automobile ignition and unlocking; anti-theft devices.

Real Life Applications

 Iris scanning forms the basis for Aadhaar Card.

 Border Patrolling in UAE.
 Iris scanning permits the passport in countries US, Canada.
 Iris scanners are employed to control access of Datacentres of Google.

Check Your Progress 3

4) Explain in detail about the image acquisition in iris scan technology.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

5) Draw the structure of Iris.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

91
Human 4.2.3 Retinal Pattern Biometrics
Identification:
Establishing
Identity-I Dr. Carleton Simon and Dr Isodore Goldstein in 1930-1935 conducted first
scientific study to describe how every individual possesses a unique Retinal
structure. Because of the differing distribution of blood vessels, they
suggested the use of the Retina to confirm the identity of an individual. Dr.
Paul Tower (1950) reported that even identical twins have a very distinct and
unique set of Retinas and has DNA Strand leading to very stable Biometric
modality and that hardly changes.

Human retina is a thin tissue composed of neural cells located in the posterior
portion of the eye. Retina is a complex structure with a network of blood
capillaries and vessels resulting into the unique feature for every person as
identical twins do not have same features. Retinal patterns are altered due to
medical morbidities otherwise it is same from birth to death. The unique
Retinal patterns of the individual is used as biometric identifier and the blood
vessels within the retina absorb light more readily than the surrounding tissue
which forms the typical smallest template for the Biometric Technology. The
principle for the Retinal scan is carried out by casting an unperceived beam
of low-energy infrared light (IR) into a person’s eye and beam traces a
standardised path on the retina as retinal blood vessels are more absorbent
than the rest of the parts of the eye, then this path is converted to computer
mode.

Figure 22: Retina Structure

Source:https://jackwestin.com/resources/mcat-content/vision/structure-and-
function-of-the-eye

92
4.2.3.1 Process of Retinal Recognition Biometrics

In comparison to Iris recognition where images are collected from a far

distance but Retinal recognition requires end user must be in close proximity
to retinal scanning device. The process has following steps:

1) Image/Signal Acquisition and Processing:

The individual eye resides on tiny receptacle to receive infrared light beam
for the illumination of retina at 3600. The duration of the process is few
minutes with precaution that the individual must remain still till one rotation
is completed. Five images are taken to create the composite image from
which the unique features are extracted.

2) Enrolments and Verification Template Creation:

Retinal patterns are not genetically controlled thus 400 unique data points are
captured from the Retina, unlike the fingerprint, where only 30-40 unique
data points are captured to create the Enrollment Template with a size of 96
bytes. The small size makes it possible to store number of templates in a
single database and same process leads to creation of Verification Template.
Due to the close proximity of numbers of factors influence the quality of
Enrollment and Verification Templates as mentioned below:

Individual cooperation is lacking

The individual’s eye and the lens distance is more.
The size of the individual’s pupil:

Figure 23: Retinal Recognition

Source: https://platform.keesingtechnologies.com/retinal-recognition-pros-and-cons/

4.2.3.2 Advantages and Disadvantages

The advantages of Retinal Recognition are:

 Due to stability and hardly ever changes over the lifetime an individual
retina is a reliable Biometric technology.
 Small sized images give quick results for the identity of the individual.
 More number of unique data points and makes error less identity.
 Retina is present within the eye thus not affected by the external
environment like fingerprints.

Disadvantages

The individuals are not comfortable to place their eye on the receptacle
with the direct infrared light.
93
Human Retinal Recognition requires high level of cooperation and motivation
Identification:
Establishing from the end user to capture high quality, raw images leading to low
Identity-I verification with a metric of 85% as compared to 100%.
The cooperation of the end user leads one to take numerous attempts and
a long time till the final result is obtained therefore there are many false
rejections.

4.2.3.3 Criteria for Effectiveness and Viability of Retinal

Recognition
Overall, the effectiveness and viability of Retinal Recognition is examined
against seven criteria as mentioned below:

Universality
Uniqueness
Permanence
Collectability
Performance
Acceptability
Resistance to circumvention

The retinal template generated is typically one of the smallest of any

biometric technology.

4.2.3.4 Similarities and Differences:

While both iris and retina scanning are ocular based biometric distinct
technologies but there are evident similarities and differences as mentioned
below:

Similarities:

 Low occurrence of false positives.

 No two individuals have the same Iris and Retina pattern thus reliable
technique.
 Identity of the individual is authenticated very fast.
 The capillaries in the iris and retina decompose too rapidly to use an
amputated eye to gain access.
Differences:
o Retinal scan measurement accuracy can be affected by disease; iris fine
texture remains remarkably stable.
o Taking a photograph is similar to iris scan from distance while for the
retinal scanning the individual has to be near like seeing a slide under
microscope.

94
o Iris scanning is more widely accepted as a commercial modality than Biometrics

retinal scanning
o Retinal scanning is invasive but not iris.

The primary function of Iris is to allow the requisite amount of light required
for proper vision. The retina is located at the back of the eye and links front
of the optic nerve to the brain therefore the function of retina is to send visual
information and cues to the brain for processing.

In other words, at Retina, image is formed that means it acts like a film of
camera.

Check Your Progress 4

6) Write a short note on retinal scanning technology.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

4.2.4 Facial Recognition Biometrics

Another Biometric Technology includes the Face Recognition a technique
that is capable of identifying or verifying a subject through an image. This
method employs measures of face and head to identify the identity of a
person through the facial biometric pattern. The first facial recognition was
carried out in 1960 based on the features of eyes, nose, etc. The 21 subjective
markers e.g. hair colour, lip thickness were employed by Goldstein and
Harmon in 1970 to automate the recognition. The standard linear algebra
techniques were applied in 1988 by Kirby and Sirovich for face recognition.
The reasons that Face Biometric is beneficial than the Biometrics are:

No physical interaction on behalf of the user is required.

Accurate with high enrolment and verification rates.
No experts as such needed for interpretation of the results.
Only biometric that allows passive identification in a one to many
environments.

4.2.4.1 Facial Recognition System

Two types of comparisons are done
Verification- It is comparison of the given individuals who are similar by the
system and then result is yes or no decision.
Identification- The given individual is compared to all the entities present in
the database thus deciphering a rank list of matches.
There are four stages in Identification technologies i.e.
i) Capture: During enrolment or verification or identification process, the
system captures a physical or behavioural sample.
95
Human ii) Extraction: The creation of template by extracting unique data.
Identification:
Establishing iii) Comparison: Compare the template with a new sample.
Identity-I
iv) Match/non-match: Matching the features extracted from new sample
and decide whether matched or not.

Match/Non
Capture Comparison Match
Extraction

Accept/
Project

Figure 24: Four Stages of Identification

Source:https://www.researchgate.net/publication/328577882_Face_Recognition_Based_on_
PCA_and_DCT_Combination_Technique/

4.2.4.2 Components of Face Recognition Systems

Enrolment Module: An automated mechanism that scans and captures a
digital or an analog image of living personal characteristics

Database: Another entity which handles compression, processing, and

storage of the captured data with stored data

Identification Module: The third interfaces with the application system

Figure 25: Components of Face Recognition

Source:https://www.researchgate.net/publication/328577882_Face_Recognition_
Based_on_PCA_and_DCT_Combination_Technique/
96
4.2.4.3 Implementation of Face Recognition System Biometrics

There are four stages in implementation process as mentioned below:

i) Image Acquisition
ii) Image Processing
iii) Distinctive Characteristic Location
iv) Template Creation And Matching

i) Image Acquisition
The static camera or video system acquires image of quality and resolution,
an essential component for verification, identification and authentication.
ii) Image Processing
The image processing steps:
 Detect the face or faces at crime scene
 Normalise and localise the face to bring the dimensions in live facial sample
 Crop the image to get ovoid facial image
 Coloured images conversion to black and white
 Initial comparisons on grey scale basis
 Align the facial sample with the templates
Distinctive Characteristic Location
The features used for matching are upper ridges of the eye sockets, areas
around the cheekbones, sides of the mouth, nose shape, and the position of
major features relative to each other and these features hardly change.
Template Creation and Matching
Processed facial images help to create enrolment template of varying size
from 100 bytes to 3K and larger templates are related to behavioural
biometrics. Matching is done 1:1 verification within facial scan system and it
is not as effective as other scans and many are returned.

Figure 26: Template Creation

Source:https://www.researchgate.net/publication/328577882_Face_Recognition_
Based_on_PCA_and_DCT_Combination_Technique

97
Human 4.2.4.4 Face Recognition Systems Working
Identification:
Establishing
Identity-I Everyone has distinguishable points on the face i.e. valleys and peaks which
are defined as nodal points in software. There are 80 nodal points on human
face that are measured by software i.e.

distance between the eyes

width of the nose
depth of the eye socket
cheekbones
jaw line
chin, etc.

For the general representation of face 70 nodal points are used to construct
the face bunch graph and an image template is matched with this face graph
in order to get the same points. The measured nodal points are given
numerical code and strings of numbers represent a face in the database called
face print.

Figure 27: Face Bunch Graph

Source:https://www.semanticscholar.org/paper/Face-Recognition-for-
Smart-Environments-Pentland

4.2.4.5 Strengths and Weaknesses of Face Biometrics

The advantages of face biometrics are as:
 convenient and has social acceptability as only picture is required.
 person is not required only image can do.
 most inexpensive biometric and has leverage for existing image.
 searching can be done on basis of static image like license and can be
performed without user cooperation.
This system has one weakness:
• cannot differentiate between identical twins.

4.2.4.6 Applications
Face recognition biometrics has replaced PIN naturally and other applications
as listed below:
Government Use:
• Law Enforcement
98 • Security/Counter terrorism
 • Immigration: Rapid progression through customs Biometrics

• Voter verification
Commercial Use:
• Residential Security: Alert homeowners of approaching personnel
• Banking using ATM: Customer’s face identified quickly with the software
• Physical access
• Face Recognition technology has now progressed and become cost
effective, reliable and highly accurate thus being applied at many
places.
4.2.5 DNA Biometrics
Behavioural or physiological traits in Biometric systems are pertaining to
individual recognition and out of which DNA profiling and fingerprinting is
the most reliable accurate method as it is intrinsically digital, and not altered
during lifetime or death. DNA profiling is known as ultimate identifier.
DNA Biometrics is defined as a specific pattern obtained from a person or
sample from the crime scene. DNA fingerprinting, is also known as DNA
typing, DNA profiling, genetic fingerprinting, genotyping, or identity testing
and it is the method of isolating and identifying variable elements within the
base-pair sequence of DNA (deoxyribonucleic acid). British geneticist Alec
Jeffreys (1984) developed the DNA Fingerprinting with a principle of
isolating and identifying variable elements within the base-pair sequence of
DNA which do not participate in functioning and highly variable called
minisatellites. Jeffrey reported that unique pattern of minisatellites are
present in every individual and only exception is identical twins from single
zygote. At present DNA fingerprinting is making usage of microsatellites of
short tandem repeats (STR) of 2-5 base pairs long.
Every individual has same chemical structure with a unique set of genes and
1% of DNA differs from person to person in sequence arrangement and it can
be understood as STOP and POTS use same letters but meaning differ. DNA
profiling is dependent on VNTR (Variable Number Tandem Repeats) which
are of two types: minisatellites and microsatellites. DNA profiling uses
repetitive sequences. VNTR loci are very similar between closely related
individuals. The analysis of variable number of tandem repeats (VNTRs) is to
detect the degree of relatedness to another sequence of oligonucleotides,
making them ideal for DNA finger printing. There are two principal families
of VNTRs:
a) Microsatellites
b) Minisatellites
Microsatellites are defined as repeating sequences of 2-6 base pairs of DNA
which are also called Simple Sequence Repeats (SSRs) or Short Tandem
Repeats (STRs). VNTR of 10-60 base pairs are named as minisatellites and
their analysis has direct application genetics, forensics, biology and DNA
fingerprinting.
99
Human 4.2.5.1 Technique of DNA Fingerprinting
Identification:
Establishing Step-1 Collection and Isolation of DNA
Identity-I
Reference Sample i.e. DNA sample of an individual of buccal swab or blood
or biological samples collected from the crime scene. Significance of DNA
analysis is that very small amount of sample is required in microliters. Alkali
and salt treatment is done on sample to get purified DNA without proteins.
Step-2 Cutting, Size and Sorting
DNA is cut into small fragments by employing Restriction Enzymes which
cut DNA at specific sites e.g. EcoR1 will cut DNA when the
sequence5’..GAATTC..3’ occurs.

Figure 28: Each Enzyme cuts at Specific Site

Source: sigmaaldrich.com/IN/en/product/sigma/r0260

Cutting of DNA yields short and long fragments which are sorted according
to size

Step-3 Separation of Fragments

Agarose gel electrophoresis is used to separate the DNA fragments on the

basis of size. The negatively charged DNA moves towards positive end and
shorter segments move faster than the longer thus DNA is sorted on the basis
of size. The electrophoresis to sort more than one sample at one time as
different samples can be loaded in the wells. Ethidium Bromide is added to
get fluorescent bands which can be visualized under UV illuminator.

Figure 29: Agarose Gel Electrophoresis and Bands on the Gel

Source:https://www.researchgate.net/publication/280232114_Isolation_and_Characterizatio
n_of_Mercury_Resistant_Bacteria_from_Fly_Ash_Sample_of_Mejia_Thermal_Power_Plant
_W_B_India_for_Application_in_Bioremediation_and_Phytoremediation/figures?lo=1

100
Step-4 Southern Blotting and Hybridisation Biometrics

A Molecular technique used to transfer DNA from gel to nylon nitrocellulose

membrane by applying blotting technique. Radioactive or coloured probes
are added to nylon sheet which sticks to only one or two specific places thus
probe is hybridised and pattern developed is DNAs fingerprint.

Step-5 DNA Fingerprint

The final DNA fingerprint is built by using several probes (5-10 or more)
simultaneously and the bar codes are used like in shops. DNA Fingerprint is
then analysed by comparing the victim and the suspect. The matching of the
profile generates strong evidence that the suspect was at the crime scene but
does not hint that he/she has committed the crime as it does not match, then
eliminated from the enquiry.

Figure 30: Steps in DNA Fingerprinting

Source: https://allaboutdnafingerprinting.weebly.com/steps-of-dna-fingerprinting.html
4.2.5.2 Applications of Fingerprinting
Diagnosis and Developing Cures for Inherited Disorders:
DNA fingerprinting is used to diagnose inherited disorders in both prenatal
and new-born babies in hospitals around the world along with the probability
of occurrence of the disease.
Biological Evidence to Identify Criminals:

DNA Fingerprinting is helpful in forensics as the biological specimens present at

the crime are important source of DNA and proves to be valuable source of
DNA of the criminal and innumerable cases are solved on this basis.
101
Human 3) Paternity Disputes: DNA Fingerprint is utilised to solve paternity
Identification:
Establishing dispute cases in court as fingerprints unprecedented with accuracy and
Identity-I reliability.

4) Personal Identification: As same DNA is present in all the body tissues

and organs thus DNA profiling is the best tool for personal identification

4.2.5.3 Advantages and Disadvantages of Fingerprinting

Advantages of DNA matching

 Accurate method of identification and has become a reliable and matured

method of identification.
 Used for medical purposes e.g. the success of bone marrow transplant is
dependent on DNA matching.
 No system performance issues.

Disadvantages of DNA matching

DNA profiling requires a physical sample for collecting biometric data

and has to be stored with appropriate environmental conditions.
DNA Samples can be prone to contamination.
It is not fully automated yet and fast unlike other biometric methods even
expensive.

Let us see the Comparison

DNA Conventional Biometrics

Requires an actual physical Uses an impression, image, or recording
sample
Not done in real-time; not all Done in real-time; automated process
stages of comparison are
automated
Does a comparison of actual Uses templates or feature extraction
samples

Check Your Progress 5

7) Draw a diagram of the procedure of DNA fingerprinting.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

102
8) How can DNA be a useful tool in the forensic applications? Biometrics

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

4.2.6 Handwriting and Voice Biometrics

Voice recognition is a crucial Biometric identifier as it recognises the speaker
as well as the speech. The measurable, physical characteristic or personal
behavioural traits are applied to verify and authenticate an individual. The
two components, Speech recognition and Speaker verification are different as
former identifies what you are saying and the later verifies who you are.

Figure 31: Speech Recognition

Source:https://www.researchgate.net/publication/334285447_Project_Report_On_AI_Speec
h_Recognition_System/

A speech recognition system assists the speaker in accomplishing what that

person wants to do and to identify the speaker then the system has to
authenticate the person and verify that he or she is in fact that person.

Figure 32: Voice Recognition Model

Source: https://developer.nvidia.com/blog/how-to-build-domain-specific-automatic-speech-
recognition-models-on-gpus/
103
Human Biometric System is the combination of speech recognition and speaker
Identification:
Establishing verification systems
Identity-I

Figure 33: Biometric System

Source: https://www.researchgate.net/

Voices have unique features like frequency and intensity which is

manipulated in the system to digitising a person's speech to produce a “voice
print”. See figure 35.

Figure 34: Voice Print

Source: https://developer.nvidia.com/blog/how-to-build-domain-specific-automatic-speech-
recognition-models-on-gpus/

In the second step these smaller segments are “captured” in a digital form and
plotted on a spectrum.

Figure 35: Voice Print Spectrum

Source: https://developer.nvidia.com/blog/how-to-build-domain-specific-automatic-speech-
recognition-models-on-gpus/
104
Third step is the Verification Process stored in a table of numbers, where the Biometrics

presence of each dominant frequency in each segment is expressed as a

binary number 0 and 1.

Figure 36: Biometric Process

Source: https://www.slideshare.net/kingsingh33/biometrics-50728903

Benefits of Voice Recognition Systems

Another layer of security for computer systems which are more efficient and
cost effective adding up no ID access cards thus saves time.

Figure 37: Increased Security Level

Source: https://www.slideshare.net/kingsingh33/biometrics-50728903
105
Human Drawbacks of Voice Recognition Biometrics
Identification:
Establishing
Identity-I Voice changes in the life time
Uses of microphones along with environmental noise
Poor phone, or cellular connections leading to Vocal variations: accents
Voice samples are converted and transferred from an analog format to a
digital format for processing by telephone devices.

Current Applications of Voice Biometrics

Security system for physical access entry into a building, in telephone

banking ATM machines and Voice verification in credit cards.
Corrections industries: inmates on parole.
Telephone companies use voice-enabling services.

The security provided by voice recognition is main benefit and further

development can be make it most successful and largest applications of
biometrics in the future to come.

Check Your Progress 6

9) Voice Recognition is becoming popular. Explain.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

4.2.6.1 Handwriting
Handwriting detection is a technique operated by Computers with the
processes to receive and interpret intelligible handwritten input from sources
e.g. paper documents, touch screen, photo graphs etc.

Way to Recognize
Handwriting

Optical Character Recognition Intelligent Word

Recognition
Fig. 38 Handwriting Detection

106
 Biometrics

Writer-1
Identification
System Writer-2

Query Sample

Database of Known Authorship

Samples

Figure 39: Working of Handwriting System (by author)

Two Types Recognition

Online Recognition
Offline Recognition

4.2.6.2 Online Recognition

On-line handwriting recognition is based on the automatic conversion of text
which is written on a special digitiser and the sensor with it picks up the pen
up and down movements known as digital ink, thus converted to letter codes.
The elements of the online recognition are:

 User can write with stylus or pen.

 A touch sensitive surface integrated with, or adjacent to, an output
display.
 The interpretation of the movements of the stylus across the writing
surface or translation of stokes into digital text is done by software.

4.2.6.3 Off-Line Recognition

Off-line handwriting recognition consists of the automatic conversion of text
in an image into letter codes by the computers for text-processing
applications regarded as static representation of handwriting. The different
styles of writing make offline process difficult. Recognition strategies are
based upon the type of data and writing and first step is Character
Recognition which is based on following criteria:

Pre-processing Methods on the data and includes the use of global

transforms (correlation, Fourier descriptors, etc.), local comparison (local
density, intersections with straight lines, variable masks, etc.); geometrical or
topological characteristics
107
Human The type of the decision algorithm relies on various statistical methods,
Identification:
Establishing neural networks, and structural matching
Identity-I
Two strategies are employed for the word recognition:

1) The Holistic Approach - recognition is globally performed on the whole

representation of words without segmentation thus an advantage of the
method.

2) The Analytical Approach - deals with several levels of representation

i.e. words are understood as sequences of smaller size units and easily
related to characters in order to make recognition independent.

Handwriting Recognition objective is to design systems for recognition of

handwriting of natural language. Methods and recognition rates depend on
the level of constraints on handwriting and characterised by the:

types of handwriting
number of scriptors
size of the vocabulary
spatial layout

4.2.6.4 Advantages and Disadvantages of Handwriting Recognition

Advantages:

Offline handwriting recognition has varied benefits like reading postal

addresses, bank check amounts, and forms. It forms the basis for digital
libraries, allowing the entry of image textual information into computers by
digitization, image restoration, and recognition methods.

Disadvantages

The disadvantage of Handwriting is that information related to data or about

the type of the input cannot be obtained. First, the text has to be separated
into characters or words to match these words in a sequence of data by usage
of algorithms.

Figure 40: Handwriting Recognition Steps

Source:https://www.researchgate.net/publication/350286160_A_Review_of_Various_Hand
writing_Recognition_Methods

108
Handwriting methodologies are difficult to operate but generate accurate Biometrics

data.

Check Your Progress 7

10) Describe offline and online handwriting systems.

……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………
……………………………………………………………………………

4.3 SUMMARY
Biometrics helps to establish a person’s identity on the basis of personal traits
by employing information technology techniques with accuracy. Biometrics
is a technique employed to measure, compare, analyse and identify the person
in terms of characteristics features like fingerprints, face, retina etc. The
personal characteristic attributes like hand geometry, colour of iris, or
behavioural characteristics etc. make every individual different thus
biometrically scan, compare and identify with following three types
modalities:

• Physiological
• Behavioural
• Both physiological and behaviour

The authentication, verification, security and accuracy has led many

applications of Biometrics in technology world like:

• Workplace access control

• Identification citizenship and immigration systems of the persons
• Secrecy and confidentiality of the information
• Forensic application in criminal Identification
• E-commerce secure transactions
• Reduction in fraud and theft
• Enforcement of law

4.4 REREFERNCES
Daugman, J.G. (1993). High confidence visual recognition of person by a test
of statistical independence. IEEE Trans. PAMI 15, 1148-1161

Hawthorn, M.R. (2009). Fingerprints: Analysis and understanding. Boca

Raton, FL: CRC Press.
109
Human Jain, A.K., Flynn, P., & Ross, A.A. (Eds.).( 2008). Handbook of biometrics.
Identification:
Establishing Springer Science+Business Media, LLC
Identity-I
Jain, A.K., Ross, A., & Nandakumar, K (2011). Introduction to biometrics: A
textbook. Berlin, Germany: Springer.

Parmar, D.N, & Mehta, B.B. (2014). Face recognition methods and
applications.

4.5 HINTS/ ANSWERS TO CHECK YOUR

PROGRESS
1) Biometrics is defined as the applicable means of identifying and
authenticating individuals in a reliable and fast way through the use of
unique biological characteristics and technologies that lead to the
acceptable way of solving the crime. The identification of persons based
on unique and definite characteristics leads to its application in forensic
science. Refer to section 4.1.3 for detail.

2) The process of comparison between the known fingerprint against

another in order to determine that it is of the same finger or palm is
known as fingerprint recognition. Refer to section 4.2.1.4 for other parts
of answer.

3) Finger Print Biometrics is defined as the method of identifying finger

print by employing the software on the basis of pattern of finger and
location of the minutiae. There are 6 patterns of fingerprints-Arch,
Tented Arch, Left Loop, Right Loop, Whorl and Twin Loop.

4) This is an important step as image acquired would lead to recognition

and the characteristics required are: Good contrast, and high quality
image and high illumination. Refer to section 4.2.2.5.

5) Iris is made up of pigmented elastic tissue with an adjustable circular

opening in the centre. Refer to section 4.2.2, Figure 14.

6) Retinal recognition requires that the end user must be in close proximity
to retinal scanning device. The process has following steps:

110
 • Image/Signal Acquisition and Processing: Biometrics

• Enrolments and Verification Template Creation.

Refer to section 4.2.3.1 for details.

7) Procedure of DNA fingerprinting involves the main steps of- Collection

and isolation of DNA, Cutting, size and sorting, Separation of fragments,
Southern blotting and hybridisation, and DNA Fingerprint. Refer to
section 4.2.5.1, Figure 31.

8) The double helical structure of the DNA present on saliva, hair follicles,
bones, blood and sperm serve as a useful tool in the forensic studies. This
can be done as the DNA from an individual’s tissue shows the same
degree of polymorphism. These polymorphic characters are inheritable
from parents to their children.

9) Benefits of voice recognition systems:

 Another layer of security for computer systems which are more

efficient and cost effective adding up no ID access cards thus saves
time.

 Security system for physical access entry into a building, in

telephone banking, ATM machines and voice verification in credit
cards.

 Correction industries: inmates on parole.

 Telephone companies use voice-enabling services.

 The security provided by voice recognition is the main benefit and

further development can make it most successful and largest
application of biometrics in future.

10) On-line handwriting recognition is based on the automatic conversion of

text which is written on a special digitiser and the sensor with it picks up
the pen up and down movements known as digital ink, thus converted to
letter codes.

Off-line handwriting recognition consists of the automatic conversion of

text in an image into letter codes by the computers for text-processing
applications regarded as static representation of handwriting.

111