Essay: An overview of the various biometric approaches

Abstract
A Biometrics system is actually a pattern recognition system that utilizes various patterns like iris, retina and biological traits like
fingerprint, voice recognition, facial geometry and hand geometry etc. What makes Biometrics really attractive is that the various
security codes like the passwords and the PIN can be interchanged between people but the physiological traits can’t be. However,
the misuse and theft of these security measures are also on the rise. To overcome from such problems we have a new solution i.e.,
ECG pattern matcher. This paper gives the overview of key biometric technologies and basic techniques involved followed by the
uses, benefits and drawbacks. The various opportunities for ECG are mentioned.
Key Words: Biometric, biometric security system, biometrics concerns, recognition methods, identification, access
control, facial recognition, fingerprint reader, voice recognition, iris/retinal recognition, vein recognition, DNA
recognition, privacy, safety.
——————————————————————–***———————————————————————-
1. INTRODUCTION
Biometrics for authentication: Biometrics traits are the
strongest link between a person and his identity as it cannot
be easily shared, lost or duplicated. Hence it is more
resistant to social engineering attacks. This type of system
requires user to be present at the time of authentication. It
can deter user from making false claims. Hence it can be
incorporated in security applications.
Law enforcement agencies in world rely on fingerprints for
criminals and forensic identification. Biometric traits deal
with access control, checking for multiple access control,
international border crossing and secure identification
documents. Each biometric trait has its own advantages and
weak points. The technique of using biometric methods for
identification can be widely applied to forensics, ATM
banking, communication security, attendance management
systems, and access control. It also plays an important role
in enhancing homeland security.
Biometric techniques involve ‘metrics’ or measurements of
some kind, rather than depending merely on familiar or
hidden methods [8].
Taxonomy of Biometric Techniques given below:
‘ Appearance – These are the physical descriptions
like color of skin or eyes, texture of hair, gender,
race, physical markings, height and weight.
‘ Social behavior -these features correspond to style
of speaking, habituated actions and visible
handicaps.
‘ Bio-dynamics – the way in which a person signs,
rhythm of speaking, keystroke dynamics,
particularly in relation to login-id and password.
‘ Natural physiography ‘These include patterns of
fingerprints, hand geometry, retina, iris, DNA,
earlobes and many more.
‘ Imposed physical characteristics like dog-tags,
collars, bracelets and anklets, bar-codes, embedded
micro-chips .[8]
2. BACKGROUND
Biometric System: Biometric system is a computer that
implements biometric algorithms and makes use of sensing,
feature extracting and matching modules. The sensing part
captures the traits, feature extraction is done to eliminate
_______________________________________________________________________________________
Volume: 03 Issue: 01 | Jan-2014, Available @ http://www.ijret.org 1
IJRET: International Journal of Research in Engineering and TechnologyeISSN: 2319-1163 | pISSN: 2321-7308
unnecessary information and the matching modules match
the traits with the references stored in the database. There
are two stages in the authentication process. They are the
enrollment and verification. Enrollment is the process of
storing the traits in the database and verification the process
of matching the extracted traitswith references in stored in
database.
Fig 2.1: Biometric System [8]
Classification of biometric identification systems:
Biometrics systems is broadly classified into two types
‘ Physical biometrics: It is based on data derived
from direct measurements of parts of the human
body. Fingerprints, iris, retina, hand geometry and
facial recognition are some of the leading
physiological biometrics.
‘ Behavioral characteristics: Identification of the
person is based on his unique behavioral
characteristics. These characteristics can be voice,
speech, signature, and rhythm of typing or an
individual’s walking style (gait). Even though the
behavioral biometrics is dependent on the actions
of an individual, it is also influenced by the
physical structure of the human body.
3. VARIOUS BIOMETRICS APPROACHES
Following section lists the various Behavioral Biometrics
approaches,
‘ Keystroke or Typing Recognition:
‘ Speaker or Voice Authentication:
A voice biometric is a numerical representation of
the sound, rhythm, and also pattern of an
individual’s voice. A voice biometric or “voice
print,” is as unique to an individual as any other
biometric methods. Voice authentication is a fairly
simple process. To register, a user records the
sample of his voice and it is stored in the
authenticating system as ‘voiceprint’. If the user
wishes to access the resource, a sample of their
voice is given to the system. A comparison is made
between the input and voiceprint to validate that
the right person is given the access to the resource.
Following section lists the various Physical Biometrics
approaches,
‘ Fingerprint Identification or Recognition:
Fingerprint technology is also known as
dactyloscopy. It is a technique of identification
where comparison is made between two instances
of fingerprints to determine whether they are from
same sources. This technique requires to compare
several features of fingerprints that are found
unique for an individual. The ridges and minutia
points are unique for an individual. The ridges can
be of three types: loop, whirl and arch .Minutia
points are specific points on fingerprints which are
critical for identification. These features of fingers
are captured as an image with help of scanner and
is enhanced and converted to a template. The
template is an encrypted biometric key or
mathematical information. The image of the
fingerprint is not stored. The algorithm cannot
convert this template back to image. Hence it is
difficult to replicate fingerprints. Scientists have
discovered that the fingerprints are inherited. It is
possible to have a belief that the members of a
family share the same fingerprint pattern
‘ Hand or Finger Geometry Recognition:
Hand geometry identifies an individual based on
the unique features of the hand. The unique
features may be length of finger, its thickness, the
distance between the finger joints and overall
structure of the bone. The system consists of a
camera which captures the image of the hand. The
necessary features are extracted, processed and
stored in the database. These stored templates can
be used later for the purpose of verification.
‘ Facial Recognition:
Face Recognition systems have the ability to
identify an individual based on the various features
on the face. Human face has several distinguishable
features called the nodal points. There are about 80
different nodal points on the face. Some of which
are the width of the nose, distance between eyes,
length of the jawline, shape of the face based on
cheek bones,depyh of the eye sockets and many
more.The positions of these nodal points are
calculated using appropriate algorithms like PCA
and LDA.The emerging trend in face recognition is
3D face recognition systems which have better
levels of accuracy when compared to the older
systems.
4. DRAWBACKS OF VARIOUS
APPROACHES
Drawbacks of various biometric systems are given below:
‘ Fingerprint:
_______________________________________________________________________________________
Volume: 03 Issue: 01 | Jan-2014, Available @ http://www.ijret.org 2
IJRET: International Journal of Research in Engineering and TechnologyeISSN: 2319-1163 | pISSN: 2321-7308
‘ Injuries, traumas, wounds or cuts can make the
fingerprint reading unidentifiable.
‘ Distortions due to grease, dirt or contamination on
finger tips have the chances of the identification to
be rejected.
‘ Today’s scanners are still not able to differentiate
between real and fake fingerprints. Fake
fingerprints can be tricked using gelatinized molds
over real finger
‘ Facial recognition:
‘ Pose variations, ageing are still the limiting factors
to identify the person using face recognition.
‘ Use of 2D scanners can’t handle pose variations
and are sensitive to variations in light and shadows.
‘ The use of 3D scanners eliminates these limits but
are expensive and have time constraints.
‘ Voice recognition:
‘ These systems are susceptible to error in the
presence of noise and external sounds. The distance
of the microphone from the user also affects the
accuracy rates.
‘ Prerecorded voice can be used for malicious
access.
‘ These systems take time to adjust to the voice of an
individual and require large amount of memory to
store voice files.
‘ Iris and retinal recognition:
‘ The distance of the person from the camera affects
the performance of the system.
‘ Errors might occur due to reflection caused by
spectacles, eyelashes or lenses.
‘ During the scanning process the person is required
to remain still.
‘ The equipment used are expensive.
‘ Hand geometry based recognition:
‘ Complications might arise when used with certain
populations. There can be a perception of bio-
hazard due to spread of potential germs.
‘ Possible changes to the shape of hands s can cause
failure of the system.
‘ Signature
‘ These kinds of systems have the limitation factor of
inconsistent signature. A person without a constant
signature may not be recognized or an individual
with muscular illness may face difficulties in
proving their identification.
‘ The quality of ink and paper may also account for
rejections.
‘ Keystroke
‘ The major drawback is low accuracy rates due to
the varying rhythms of typing. These variations
may be due to injury, fatigue, distractions, and
mood or due to the side effects of drugs,
medications or consumption of alcohol.
‘ DNA recognition:
‘ As it is a very new technology, may not be very
accurate among close relatives and hence less
popular in public.
‘ Requires time to process and establish identity.
‘ Requires expensive equipment for processing and
analyzing the samples.
.
5. ECG AS BIOMETRIC TRAIT
ECG based biometric is a recent topic for research. The
ECG record is a graphical record of electrical impulses of
heart. Electrical activity of the heart is represented by ECG
signal. Capturing the ECG signal without the cooperation
from the person is a difficult task. Hence cannot be copied
easily to provide fake identity. One of the most important
strengths of ECG is that it accounts for the vitality of a
person. Hence it requires that the person has to be
necessarily present at the time of authentication.
All these factors assure better security when compared to
other biometric traits which were discussed earlier. Hence
ECG can become the most promising biometrics in the near
future.
Description of ECG waveforms: The ECG is characterized
by five different segments of the waveform: the P, Q, R, S
and T Waves. [9] The analysis of each of the segments gives
information about the different events of the Cardiac cycle
as in fig 5.1:
P-wave: It is a small low-voltage deflection caused due to
the atrial depolarization.
QRS complex: The largest-amplitude portion of ECG,
caused due to the currents generated due to the ventricular
depolarization.It is made of thre closely related waves Q-
wave,R-wave and S wave.
T-wave: It represents ventricular repolarization.
P-Q interval:It is the time interval between the beginning of
the P wave and the beginning of the QRS complex.
_______________________________________________________________________________________
Volume: 03 Issue: 01 | Jan-2014, Available @ http://www.ijret.org 3
IJRET: International Journal of Research in Engineering and TechnologyeISSN: 2319-1163 | pISSN: 2321-7308
Features that make ECG unique
ECG is unique as the morphology and the amplitudes of the
cardiac cycle are dependent on the size, shape and position
of the heart. The most variable factor of heart is the
heartbeat. The normal heart beat of a person is around 60-80
beats/min. This may rise to 200beats/min under pressure or
excitement. This variation may reduce the diastole duration
and ventricular depolarization. It may also cause the
attenuation of the R wave amplitude. But the duration of
QRS complex does not vary significantly.
Generally the raw ECG signal is contaminated by noise and
hence requires preprocessing to remove the disturbances.
Once preprocessing is done, Feature has to be extracted for
the purpose of authentication. The features that can be
extracted are:
Angle features: The angle between PQR, QRS and RST in
the ECG signal can be used.
Interval features: The time interval between two R peaks can
also be used as one of the features. The others are the time
difference between a peak and a valley or between two
peaks or two valleys.
Amplitude features: The difference between the amplitudes
of two peaks or two valleys or one peak and a valley can be
the features extracted for authentication.
The feature to be extracted has to be selected carefully based
on feasibility, consistency and accuracy rates. The angle and
interval features are not constant .This because of the fact of
the heart beat rate changes from a child to an adult. Hence
these features are not suitable. But the amplitude features do
not change with age and remain almost constant. Hence the
amplitude features may be more suitable for the purpose of
authentication.
6. APPLICATIONS OF ECG AS A
BIOMETRIC TRAIT
We have listed following areas where we can have an
authentication process via. ECG.
‘ In attendance management systems usually ID
‘ In e-voting machines, ECG can be used as key
‘ In locking systems present in mobile phones,
‘ It can be used in the field of telemedicine to
‘ Used in application involving financial
‘ Used as digital signature to file income tax
‘ May be used in Aadhar cards. Aadhar is 12
‘ Can be used as factor of identification of a
‘ Gain access control over data files and
cards are used for validating the presence of an
individual. It is possible for a person to forget
his ID at home but very unlikely to forget his
heart at home. It is also difficult to make proxy
attendance using ECG.
factor of single vote. A person can cast single
vote. Addition of fake votes or multiple votes
can easily be identified and rejected.
house doors, bank lockers, vehicles etc.
monitor a patient’s health over long distances
using ECG as identifying factor.
transactions like net banking, ATM systems,
online shopping systems etc.
returns, eProcurement, eTendering, sending
and receiving encrypted mails etc.
digits unique ID issued to every citizen of
India. This single source of truth will help in
financial inclusions, with deeper penetration of
financial institutions and smooth error free
distributions of governmental schemes. It card
has fingerprints and irises captured. It is likely
to fake these biometric factors. But if ECG
based identification is used it is difficult to
clone the uniqueness of a person.
person in international border crossing and
henceforth avoid the illegal trespassing,
smuggling activities and other anti-social
activities.
computer programs which contain personal
and crucial information which is of importance
to the military, investigations bureaus,
revenue, defense and other government
organizations. It is very important to maintain
secrecy for these agencies.
fig 5.1
7. CONCLUSION
This paper has evaluated the purpose of using ECG
enabled biometric authentication system. Unlike
conventional biometrics that is neither secure nor
robust enough against falsification, ECG is
inherited to an individual is highly secure and
impossible to be forged. The most important
feature of ECG is its real-time feature of vitality.
This ensures that ECG cannot be acquired from a
_______________________________________________________________________________________
Volume: 03 Issue: 01 | Jan-2014, Available @ http://www.ijret.org 4
IJRET: International Journal of Research in Engineering and TechnologyeISSN: 2319-1163 | pISSN: 2321-7308
dead person and compulsorily requires the presence
of the person at the time of authentication. ECG is
the most promising biometric feature due to its
uniqueness, universality and acceptability. The
characteristic of permanence is also a key factor for
it to emerge as an excellent biometric system in the
near future
Reference:
[1] [Drawbacks Of Various Approaches] Found at,
http://biometrics.pbworks.com/w/page/14811349/Advantage
s%20and%20disadvantages%20of%20technologies
[2] [Drawbacks Of Various Approaches] Found at,
https://sites.google.com/a/cortland.edu/biometrics/disadvant
ages
[3] ] B Rodger Jamieson, Ph.D., CA, Greg Stephens and
Santhosh Kumar ‘Fingerprint Identification: An Aid to the
Authentication Process’, Information Systems Audit and
Control Association.
[4]NicosMaglaveras, TelemachosStamkopoulos,
KonstantinosDiamantaras, Costas Pappas, Michael Strintzis
‘ECG pattern recognition and classi??cation using non-
linear transformations and neural networks: A review’,
International Journal of Medical Informatics 52 (1998) 191-
208
[5]Fahim Sufi, Ibrahim Khalil, and Jiankun Hu, ‘ECG-
Based Authentication’, Chapter 17, Handbook of
Information and Communication Security (Eds.)
[6][Biometric authentication] found at,
http://www.scholarpedia.org/article/Biometric_authenticatio
n
[7]Abhishek Kumar Sinha, ‘Financial Transactions get
personalized and secure with biometrics’, Digital
Transformation.
[8]P. SASIKALA and Dr. R.S.D. WAHIDABANU,
‘Identification of Individuals using Electrocardiogram’,
IJCSNS International Journal of Computer Science and
Network Security, VOL.10 No.12, December 2010
[9]Dr.Neil Townsend, ‘Medical Electronics’,Michaelmas
Term 2001 found at pno 7
[10]Steven A. Israel, John M. Irvine, Andrew Cheng,
MarkD.Wiederhold,Brenda K. Wiederhold ‘ECGto identify
individuals’,PatternRecognition,The journal of pattern
recognition society.