Human Recognition through

Human Recognition through
Iris Image Processing

Mr. Swanirbhar Majumder

Department of Electronics & Comm. Engg

North Eastern Regional Institute of Science &
Technology (NERIST)

(Deemed University)

Mr. Saurabh Pal

Department of Electronic Instrumentation Engg

Haldia Institute of Technology (HIT)

Under WBUT (West Bengal University of Technology)

Introduction

Iris

recognition

is

one

of

the

most

secure

biometric

technologies
.

Although

facial

recognition,

voice

recognition,

and

hand

geometry

are

nonintrusive

and

have

value

for

certain

applications,

all

of

these

biometric

methods

are

relatively

unreliable

compared

with

iris

recognition

Iris Facts:


The

iris

is

the

colored

part

of

the

eye
;

it

lies

behind

the

cornea

and

in

front

of

the

lens

and

is

protected

by

the

eyelid
.



The

iris

is

formed

of

a

trabecular

meshwork

(elastic

connective

tissue),

layers

of

pigment,

muscle,

and

ligaments,

and

it

controls

the

amount

of

light

that

enters

the

eye

by

allowing

the

pupil

to

dilate
.



Color

is

not

used

in

iris
-
recognition

technology
.

Instead,

the

other

visible

features
—
the

connective

tissue,

cilia,

contraction

furrows,

crypts,

rings,

and

corona

distinguish

one

iris

from

another
.

Iris Facts: (Continued)


By

the

time

a

person

is

about

eight

months

old,

the

structures

of

the

iris

are

complete,

and

they

do

not

change

in

later

life
.

The

iris

cannot

be

surgically

altered

without

damage

to

a

person’s

vision



The

purpose

of

iris

recognition

products

is

to

provide

real
-
time,

high
-
confidence

recognition

of

a

person’s

identity


Iris Facts: (Continued)


The

statistical

probability

that

two

irises

would

ever

be

exactly

the

same

is

estimated

at

1

in

10
72
.

Iris

recognition

is

statistically

more

accurate

than

DNA

testing
.


No

two

irises

are

alike,

not

even

among

twins
.

In

fact,

left

and

right

irises

of

one

individual

are

not

identical


Irises

(pigmented,

round,

contractile

membranes

in

the

eyes)

are

unique

in

each

individual


Overview:


The

Iris

as

a

Biometrics
:

The

iris

is

an

overt

body

that

is

available

for

remote

assessment

with

the

aid

of

a

machine

vision

system

to

do

automated

iris

recognition
.


Iris

recognition

technology

combines

computer

vision,

pattern

recognition,

statistical

inference,

and

optics
.



The

spatial

patterns

that

are

apparent

in

the

human

iris

are

highly

distinctive

to

an

individual
.

Advantages:

Its suitability as an exceptionally accurate biometric derives from its:


extremely data
-
rich physical structure


genetic independence
—

no two eyes are the same


patterns apparently stable throughout life


physical protection by a transparent window (the cornea), highly
protected by internal organ of the eye


externally visible, so noninvasive
—

patterns imaged from a
distance

Disadvantages:

The disadvantages to use iris as a biometric measurement are:

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


Moving target


Located behind a curved, wet, reflecting surface


Obscured by eyelashes, lenses, reflections


Partially occluded by eyelids, often drooping


Deforms non
-
elastically as pupil changes size

RECOGNITION PROCESS:

Steps:

1.
Image

capture
:

Iris

recognition

begins

with

a

video

picture

of

the

eye

and

iris

within

it
.

A

monochrome

CCD

is

the

type

of

camera

used
.

2.
Image

acquisition
:

Since

iris

is

small

in

size

and

dark

in

color,

it

is

difficult

to

acquire

good

images

for

analysis

using

the

standard

CCD

camera

and

ordinary

lighting
.

Image

acquisition

provides

iris

image

of

sufficiently

high

quality
.

We

have

so

far

got

the

eye

image

after

image

acquisition

from

the

camera
.

Steps: (Continued)

3.
Preprocessing
:

It

is

done

to

remove

some

irrelevant

parts

(e
.
g
.

eyelid,

pupil

etc
.
)

from

the

image
.

Detected

the

iris

after

preprocessing
.

Applying

the

MBIR

[Minimum

Bounding

Isothetic

Rectangle

(Square)]

technique

to

get

the

iris

as

for

normalization

by

removing

the

unnecessary

parts
.

Here

we

apply

histogram

equalization

and

get

the

resultant

MBIR

image

Steps: (Continued)

4.
Database

creation
:

A

database

of

the

normalized

iris

is

created

so

that

whenever

an

iris

image

is

got

the

normalized

pattern

of

that

iris

can

be

checked

to

recognize

the

concerned

person
.

Here

we

use

canny’s

edge

detection

technique

to

get

the

iris

pattern

from

the

MBIR
-
ed

image
.

For

each

person

though

we

had

20

images

of

each

eye,

but

here

presently

we

used

5

edge

detected

MBIR

images

and

compare

the

intra

class

correlation

coefficient
.

The

lowest

of

the

correlation

coefficients

was

taken

with

5

percent

tolerance

to

be

the

optimum

correlation

coefficient
.

Then

we

store

in

memory

the

optimum

correlation

coefficient

along

with

any

one

of

the

edge

detected

binary

MBIR

images
.

.

This

we

carry

on

for

25

persons
.

All

these

are

done

in

.
mat

format

of

MATLAB®
.

Thus

we

develop

a

database

of

25

concerned

persons
.

Steps: (Continued)

5.
Matching
:

The

matching

process

includes

providing

of

an

eye

image

as

input,

from

which

we

get

the

last

Fig

as

output

after

all

computations

and

detections
.

Then

the

output

image’s

logical

or

binary

format

is

correlated

with

all

the

25

databases

and

the

correlation

coefficient

value

using

equation

is

compared

with

the

optimum

value

stored

in

the

database
.

If

the

correlation

coefficient

is

greater

than

the

value

in

the

database

than

we

take

it

for

MATCHED

with

that

specific

person

and

recognize

him/she

else

if

the

input

image

does

not

provide

optimum

correlation

then

it

is

taken

for

granted

that

the

input

image

is

of

an

unrecognized

person
.

RESULTS:



Created

the

database

of

25

images

with

their

respective

optimum

correlation

coefficients


Then

checked

these

for

10

images

each

for

25

persons

and

other

25

persons

whose

normalized

images

were

not

present

in

the

database

whose

correlation

coefficients

are

tabulated

as

in

Table

I
.



Then

we

used

250
(
25
X
10
)

for

match

and

250
(
25
X
10
)

for

non
-
match

thus

total

of

500

input

images

provided

we

got

227

matches

and

273

non
-
matches

of

which

we

got

only

two

false

matches

and

25

false

non
-
matches
.

The

details

are

as

tabulated

in

Table

II
.



RESULTS: (Continued)


The

histogram

plot

of

inter

and

intra

matching

of

initial

125

images

(
25
X
5
)

are

as

in

Fig
.

From

which

we

can

see

that

the

out

of

15500

data


CONCLUSION:


We

have

achieved

success

till

this

point

with

around

5
%

error,

the

main

task

is

still

left

as

our

technique

is

offline

and

non

adaptive
.


We

still

have

the

problem

of

normalizing

images

that

are

taken

with

the

person

wearing

glasses

or

contact

lens

or

for

images

where

the

eyelashes

cover

the

iris

region
.



We

are

planning

to

add

two

dimensional

Gabor

wavelets

to

filter

the

images

to

get

better

correlation

coefficients

to

avoid

further

false

matches

or

false

non

match

and

decrease

the

error

percentage
.


We

can

say

that

as

we

are

storing

the

images

in

logical

or

binary

form

the

representation

of

each

pixel

is

either

0

or

1

so

we

can

store

8

pixels

in

a

byte

whereas

if

we

had

kept

the

traditional

format

of

unsigned

integers

of

8

bits

e

would

require

a

byte

for

each

pixel
.


ACKNOLEDGEMENT:


We

would

mainly

like

to

thank

Mr
.

Soumyadip

Rakshit,

Researcher

in

Biometric

Signal

Processing

of

Signal

and

Image

Processing

Group

(SIPG),

Department

of

Electronic

and

Electrical

Engineering

at

the

University

of

Bath

who

provided

us

the

database

created

by

them

for

iris

image

processing
.

The

database

consisted

of

20

images

of

both

eyes

of

50

persons

amounting

to

2000

images
.



Along

with

that

we

would

also

like

to

thank

Prof
.

B
.

N
.

Chatterji

who

actually

motivated

us

to

work

in

this

field

and

in

this

context

using

MBIR

technique
.

REFERENCES:

[
1
]

S
.

Sanderson,

J
.

Erbetta
.

Authentication

for

secure

environments

based

on

iris

scanning

technology
.

IEE

Colloquium

on

Visual

Biometrics
,

2000
.


[
2
]

J
.

Daugman
.

How

iris

recognition

works
.

Proceedings

of

2002

International

Conference

on

Image

Processing,

Vol
.

1
,

2002
.

[
3
]

J
.

Daugman
.

Biometric

personal

identification

system

based

on

iris

analysis
.

United

States

Patent,

Patent

Number
:

5
,
291
,
560
,

1994
.

[
4
]

R
.

Wildes
.

Iris

recognition
:

an

emerging

biometric

technology
.

Proceedings

of

the

IEEE
,

Vol
.

85
,

No
.

9
,

1997
.


[
5
]

R
.

Wildes,

J
.

Asmuth,

G
.

Green,

S
.

Hsu,

R
.

Kolczynski,

J
.

Matey,

S
.

McBride
.

A

system

for

automated

iris

recognition
.

Proceedings

IEEE

Workshop

on

Applications

of

Computer

Vision
,

Sarasota,

FL,

pp
.

121
-
128
,

1994
.

REFERENCES: (Continued)

[
6
]

W
.

Boles,

B
.

Boashash
.

A

human

identification

technique

using

images

of

the

iris

and

wavelet

transform
.

IEEE

Transactions

on

Signal

Processing
,

Vol
.

46
,

No
.

4
,

1998
.


[
7
]

S
.

Lim,

K
.

Lee,

O
.

Byeon,

T
.

Kim
.

Efficient

iris

recognition

through

improvement

of

feature

vector

and

classifier
.

ETRI

Journal
,

Vol
.

23
,

No
.

2
,

Korea,

2001
.


[
8
]

S
.

Noh,

K
.

Pae,

C
.

Lee,

J
.

Kim
.

Multiresolution

independent

component

analysis

for

iris

identification
.

The

2002

International

Technical

Conference

on

Circuits/Systems,

Computers

and

Communications
,

Phuket,

Thailand,

2002
.

[
9
]

Libor

Masek,

School

of

Computer

Science

and

Software

Engineering,

The

University

of

Western

Australia,
2003

[
10
]

S
.

Rakshit

and

others

of

Signal

and

Image

Processing

Group

(SIPG),

Department

of

Electronic

and

Electrical

Engineering

at

the

University

of

Bath
,

http
:
//www
.
bath
.
ac
.
uk

THANK YOU!