A survey of image-based

licoricebedsSécurité

22 févr. 2014 (il y a 3 années et 1 mois)

84 vue(s)

A survey of image
-
based
biometric identification methods:

Face, finger print, iris, and others


Presented by: David Lin

ECE738 Presentation of Project Survey

© 2003 by David Lin

2

Outline


Problems and motivations


Different identification methods


Face Recognition


Fingerprints


Iris Recognition


Hand Geometry


Others


Summary and Conclusions


© 2003 by David Lin

3

Problems


Security has always been an important concern
to many people. Such as banks, industrial,
military systems, and personal information.


Traditional security and identification are base
on things that can be easily breached.
Knowledge based or token based.


Not unique, can be duplicated, e.g. Passwords
and ID cards.


© 2003 by David Lin

4

Biometrics System


Identity verification of living, human
individuals based on physiological and
behavioral characteristics.


“Something you are or you do”


In general, biometric system is not easily
duplicated and unique to each individuals


© 2003 by David Lin

5

Biometrics System


What should we look for in Biometrics systems?


Universality, which means that each person should
have the characteristic


Uniqueness, which indicates that no two persons
should be the same in terms of the characteristic


Permanence, which means that the characteristic
should not be changeable


Collectability, which indicates that the characteristic
can be measured quantitatively


© 2003 by David Lin

6

Face Recognition


Techniques such as, Eigenfaces,
geometry representation, Gabor wavelet
transform, Karhunen
-
Loeve, etc.


Acquisitions
-

frontal view, half profile,
profile view.


Affected by facial beard, glasses, hair style,
age.



© 2003 by David Lin

7

Fingerprints


Most of the existing
systems uses
“minutiae” in a
fingerprint image for
matching.


Minutiae are the
details in the
fingerprint ridges,
ridge endings and
bifurcations.



Endings

Bifurcations

© 2003 by David Lin

8

Fingerprints











1
1
1
1
0
1
1
1
1
Extraction Filter

1 = ending

2 = ridge

3 = bifurcation

© 2003 by David Lin

9

Iris Recognition


The highly randomized appearance of the iris
makes its use as a biometric well recognized. Its
suitability as an exceptionally accurate biometric
derives from its,


extremely data
-
rich physical structure,


genetic independence, no two eyes are the same


stability over time


physical protection by a transparent window (the
cornea) that does not inhibit external viewability.

© 2003 by David Lin

10

Iris Recognition


Daugman Method, zero
-
crossing 1D wavelet
transform, multi
-
channel Gabor filtering


Most of them uses Gabor wavelets filter


Iris code is calculated using circular bands that
have been adjusted to conform to the iris and
pupil boundaries.


Eyelashes or the eyelid obscure part of the grid
might influence system operations

© 2003 by David Lin

11

Multi
-
channel Gabor filtering

Extracted block is 512 x 64 pixels


Daugman Method

Eight circular band

512
-
byte iris code

© 2003 by David Lin

12

Hand Geometry

Different views of the prototype designed: (a) Platform and camera,

(b) placement of the user's hand, and (c) photograph taken.

Measurements



Widths



Heights



Deviations



Angles


Classifiers



Euclidean Distance



Hamming Distance



Gaussian Mixture Models


GMM shows the

best result

© 2003 by David Lin

13

Hand Vein Patterns


Hand vein pattern is
distinctive for various
individuals.


The veins under the skin
absorb infrared light and thus
have a darker pattern on the
image of the hand taken by
an infrared camera.


One system is manufactured
by British Technology Group
is called Veincheck and uses
a template with the size of 50
bytes.

Back of the hand

© 2003 by David Lin

14

Retinal Patterns


Uses the vascular patterns
of the retina of the eye.


In healthy individuals, the
vascular pattern in the
retina does not change
over the course of an
individual's life.


The patterns are scanned
using a low
-
intensity (e.g.
near
-
infrared) light source.

© 2003 by David Lin

15

Retinal Patterns


The main drawback of the retina scan is its
intrusiveness. A laser light must be
directed through the cornea of the eye.


Operation of the retina scanner is not easy.


The size of the eye signature template is
96 bytes.

© 2003 by David Lin

16

Signature


Uses the dynamic analysis of a signature to
authenticate a person.


Measuring dynamic features such as speed,
pressure and angle used when a person signs
a standard, recorded pattern (e.g. autograph).

Captured using a tablet


One focus for this technology
has been e
-
business
applications and other
applications where a signature
is an already accepted method
of personal authentication.


© 2003 by David Lin

17

Summary & Conclusions

Ease of
use

Error
incidence

Accuracy

User
acceptance

Required
security
level

Long
-
term
stability

Fingerprint

High

Dryness, dirt

High

Medium

High

High

Hand
Geometry

High

Hand injury,
age

High

Medium

Medium

Medium

Iris

Medium

Poor
Lighting

Very High

Medium

Very High

High

Retina

Low

Glasses

Very High

Medium

High

High

Signature

High

Changing
signatures

High

Very high

Medium

Medium

Face

Medium

Lighting,
age, hair,
glasses

High

Medium

Medium

Medium

© 2003 by David Lin

18

Summary & Conclusions


By combining two or more individual biometric systems cheaper
and reliable security can be obtained.