Biometrics

highpitchedteamΑσφάλεια

30 Νοε 2013 (πριν από 3 χρόνια και 9 μήνες)

92 εμφανίσεις

BIOMETRICS



Karthiknathan

Srinivasan



Sanchit

Aggarwal


Overview


What is Biometrics?



Measures



Biometric System



Modes of Operation



Modules



Types of Biometric Recognition



Applications



Advantages/Disadvantages

What is Biometrics?

Methods of identifying a person based on

Physiological or Behavioral characteristic.




Physiological
-

Hand or finger images, facial
characteristic, speak verification, iris recognition.



Behavioral
-

Dynamic Signature Verification and
Keystroke Dynamics.


What Biological Measures Qualify to be a
Biometric


Universality
-

Each person should have the
characteristic.



Distinctiveness
-

Two persons should be different
in terms of characteristics.



Permanence
-

Characteristic should be invariant
of time.



Collectability
-

Characteristic should be measured
Quantitatively.

Biometric Systems

A biometric system is a pattern recognition system

that operates by


o
Acquiring Biometric data from an Individual.


o
Extracting Feature Set from the Data.


o
Comparing the Feature Set with the Template in
the Database.


Operation Modes Of Biometrics

There are two modes of operation.


o
Verification Mode

o
Identification Mode.



Depending on the Application Context, Biometric
System can work either on Verification Mode or in
Identification Mode.


Block Diagram of Enrollment, Verification, Identification Phase




J
A
IN
et
al.
:
A
N
IN
T
R
O
D
U
C
T
I
O
N
T
O
B
I
O
M
E
T
R
I
C
R
E
C
O
G
N
I
T
I
O
N
5
Fig.
1.
Block
diagrams
of
enrollment,
v
erif
ication,
and
identif
ication
tasks
are
sho
wn
using
the
four
main
modules
of
a
biometric
system,
i.e.,
senso
r
,
feature
e
xtraction,
matcher
,
and
system
database.
In
such
a
system,
an
indi
vidual
who
desires
to
be
recog-
nized
claims
an
identity
,
usually
via
a
personal
identif
i-
cation
number
(PIN),
a
user
name,
or
a
smart
card,
and
the
system
conducts
a
one-to-one
comparison
to
determine
whether
the
claim
is
true
or
not
(e.g.,

Does
this
biometric
data
belong
to
Bob?

).
Identity
v
erif
ication
is
typically
used
for
positive
r
eco
gnition
,
where
the
aim
is
to
pre
v
ent
multiple
people
from
using
the
same
identity
[26].

In
the
identif
ication
mode,
the
system
recognizes
an
indi-
vidual
by
searching
the
templates
of
all
the
users
in
the
database
for
a
match.
Therefore,
the
system
conducts
a
one-to-man
y
comparison
to
establish
an
indi
vidual

s
iden-
tity
(or
f
ails
if
the
subject
is
not
enrolled
in
the
system
data-
base)
without
the
subject
ha
ving
to
claim
an
identity
(e.g.,

Whose
biometric
data
is
this?

).
Identif
ication
is
a
crit-
ical
component
in
ne
gative
r
eco
gnition
applications
where
the
system
establishes
whether
the
person
is
who
she
(im-
plicitly
or
e
xplicitly)
denies
to
be.
The
purpose
of
ne
g
a-
ti
v
e
recognition
is
to
pre
v
ent
a
single
person
from
using
multiple
identities
[26].
Identif
ication
may
also
be
used
in
positi
v
e
recognition
for
con
v
enience
(the
user
is
not
re-
quired
to
claim
an
identity).
While
traditional
methods
of
personal
recognition
such
as
passw
ords,
PINs,
k
e
ys,
and
tok
ens
may
w
ork
for
positi
v
e
recognition,
ne
g
ati
v
e
recog-
nition
can
only
be
established
through
biometrics.
Throughout
this
paper
,
we
will
use
the
generic
term
r
eco
gni-
tion
where
we
do
not
wish
to
mak
e
a
distinction
between
v
eri-
f
ication
and
identif
ication.
The
block
diagrams
of
a
v
erif
ication
system
and
an
identif
ication
system
are
depicted
in
Fig.
1;
user
enrollment,
which
is
common
to
both
of
the
tasks,
is
also
graph-
ically
illustrated.
The
v
erif
ication
problem
may
be
formally
posed
as
follo
ws:
gi
v
en
an
input
feature
v
ector
(e
xtracted
from
the
biometric
data)
and
a
claimed
identity
,
determine
if
(
)
belongs
to
class
or
,
where
indicates
that
the
claim
is
true
(a
gen-
uine
user)
and
indicates
that
the
claim
is
f
alse
(an
impostor).
T
ypically
,
is
matched
ag
ainst
,
the
biometric
template
corresponding
to
user
,
to
determine
its
cate
gory
.
Thus
if
otherwise
where
is
the
function
that
measures
the
similarity
between
feature
v
ectors
and
,
and
is
a
predef
ined
thr
eshold
.
The
v
alue
is
termed
as
a
similarity
or
matc
hing
scor
e
be-
tween
the
biometric
measurements
of
the
user
and
the
claimed
identity
.
Therefore,
e
v
ery
claimed
identity
is
classif
ied
into
Operational Modes Contd.


In
Verification mode
, the system validates the
person’s identity by comparing the captured
biometric data with the template stored in the
database. This template is stored in the Enrollment
phase.



In
Identification mode
the system identifies the
person by searching the templates of all users in the
database for a match. One to many Comparison.

Modules needed to build a Biometric System



Sensor module



Feature Extraction module



Matcher Module



System Database Module




1.
Sensor Module
-

It captures the Biometric data of an
Individual. An example can be a
Fingerprint Sensor
.


2.
Feature Extraction Module
-

Here the obtained
biometric data of an Individual is processed to extract
features. Example can be the
Local ridge feature
extraction

from a Fingerprint.


3.
Matcher Module
-

Here the features extracted during
the above phase are matched against the templates
s
tored
in the database.


4.
System Database Module
-

Used to Store Biometric
templates of the users enrolled. The enrollment module is
responsible for Enrolling Individuals to the database.

Types of Biometric Recognition

Common Techni ques



Fingerprint Recognition



Face Recognition



Voice Recognition



Iris Recognition



Hand Geometry



Signature Verification

Other Techni ques



Keystroke



Ear Geometry



Lip Motion



Thermograms



Retina Recognition

Fingerprint Recognition



Taking

an

image

of

a

person’s

fingertips

and

storing

the

characteristics
.




Includes pattern matching


o

Ridges

o

Whorls

o

Arches

o

Furrows


Iris Recognition


Camera technology



Infrared illumination



Mathematical
-
pattern
recognition techniques



Facial Recognition


Recording

face

images

through

a

digital

video

camera
.



Analyzing

facial

characteristics

like

the

distance

between

eyes,

nose,

mouth

and

jaw

edges
.


Applications


ATMs



Computer Login



Online Banking



National Security



Elections



Criminal Investigation



Identification of missing people



Advantages


Easy to maintain



More robust than ID Cards, Passwords, PIN numbers,
etc.



Cannot be stolen or forgotten



Single biometric protection for multiple logins




Disadvantages


It can be very expensive



The pattern matching might be inaccurate due to
environmental conditions



The stored biometric data might be vulnerable to
malicious attacks



Reproduction of biometric data by other people

QUESTIONS?