Biometric Technologies

1
Biometric Technologies
Team 3 Steven Golikov Barbara Edington Melanie
Johnson Bashir Amhed Borming Chiang
2
Introduction to Biometrics

• History of Biometrics
• Biometrics is the study of biological data
• Biometrics has a very long tradition
• The Egyptians used the length of a persons
  forearm to determine their identification for
  wage payment
• There are many different biometrics used for
  identification
• Fingerprints
• Eye Retinal or Iris
• Facial Recognition
• Voice
• Signature
• Dental
• DNA
• www.biometricscatalog.org

3
Iris and Retinal Scanning

• The Basics of Human Eye
• Rationales
• How do the Technologies Work?
• Applications
• Performance Metrics
• Pros and Cons
• Commercial Products

4
The Basics of Human Eye

• Iris
• The Shutter of our biological camera
• The plainly visible colored ring underneath
  cornea.
• Iris surrounds the pupil
• A muscular structure which controls the amount of
  light entering into the eye, and it has very
  intricate details such as colors, striations,
  pits and furrows.

• Retina
• The film of the camera
• Located in the back of the eye where the Optic
  nerve connects.
• The blood vessels pattern in the retina are
  unique to each individual.

Source http//www.stlukeseye.com/Anatomy.asp
5
Rationales for Use
Iris contains intricate details such as
striations, pits and furrows. Two Iriss are not
alike. There is no detailed correlation between
the patterns of identical twins or even between
the left and right eye of the same
individual. Expressed by the Individuals
Phenotype, Not Genotype
Iris Collage
Retinal Image - Twin 1
Retinal Image -Twin 2
The patterns of blood vessels in the retina is
extremely unique to individuals. There is no
detailed correlation between the patterns of
identical twins or even between the left and
right eye of the same individual.
Source of images http//www.cl.cam.ac.uk/users/jg
d1000/
6
How does Iris Recognition Works?

• A picture of the eye is taken from within 1lt
  meter distance and Iris portion is extracted
• An Iris code of 512 Bytes is generated using
  functions called 2-D wavelets. This code is
  unique to one eye of one individual.
• Iris code is then compared to other Iris codes
  that are stored in the database

Source of images http//www.iridiantech.com/
7
History of Iris Recognition

• 1936 Idea proposed by ophthalmologist Frank
  Burch
• 1949 - The idea documented in an ophthalmology
  textbook by James Doggarts
• 1980's - The idea had appeared in James Bond
  films, but it still remained science fiction and
  conjecture.
• 1987 - two ophthalmologists, Aran Safir and
  Leonard Flom, patented this idea
• 1989 - John Daugman (then teaching at Harvard
  University) try to create actual algorithms for
  iris recognition
• John Daugman algorithms patented in 1994, are the
  basis for all current iris recognition systems
  and products

8
Commercial Applications -Iris

• The major applications of this technology so far
  have been
• Aviation security and controlling access to
  restricted areas at airports
• London Heathrow, Amsterdam Schiphol, Frankfurt,
  Athens, and several Canadian airports,
  Charlotte/Douglas International Airport in North
  Carolina
• Database access and computer login
• Access to buildings and homes
• Hospital settings, including mother-infant
  pairing in maternity wards
• Border control "watch list" database searching at
  border crossings
• On the Pakistan Afghanistan border, the United
  Nations High Commission for Refugees uses these
  algorithms for anonymous identification of
  returning Afghan refugees receiving cash grants
  at voluntary repatriation centres
• Other law enforcement agency programs such Jail
  Security
• Prisoner Identification 1994 - Lancaster
  County Prison in Pennsylvania became the first
  correctional facility to employ the technology.

9
Performance Comparison
Method Coded Pattern Misidentification rate Security Applications
Iris/Retinal Recognition Iris code / Blood vessel pattern 1/1,200,000 High High-security facilities
Fingerprinting Fingerprints 1/1,000 Medium Universal
Hand Shape Size, length and thickness of hands 1/700 Low Low-security facilities
Facial Recognition Outline, shape and distribution of eyes and nose 1/100 Low Low-security facilities
Signature Shape of letters, writing order, pen pressure 1/100 Low Low-security facilities
Voiceprinting Voice characteristics 1/30 Low Telephone service
Source AIM Japan, Automatic Identification
Seminar, Sept.14, 2001
10
Pros and Cons

• Iris Scanning
• The uniqueness of Irises, even between the left
  and right eye of the same person, makes iris
  scanning very powerful for identification
  purposes.
• The likelihood of a false positive is extremely
  low and its relative speed and ease of use make
  it a great potential biometric.
• The only drawbacks are the potential difficulty
  in getting someone to hold their head in the
  right spot for the scan if they are not doing the
  scan willingly.

• Retina Scanning
• Retina scan devices are probably the most
  accurate biometric available today. The
  continuity of the retinal pattern throughout life
  and the difficulty in fooling such a device also
  make it a great long-term, high-security option.
• The high cost of the proprietary hardware as well
  as the inability to evolve easily with new
  technology make retinal scan devices a bad fit
  for most situations.
• It also has the stigma of consumer's thinking it
  is potentially harmful to the eye, and in
  general, not easy to use.

11
Commercial Products and Vendors
Iris scanning (very accurate, expensive) Argus Solutions (Australia) http//www.argus-solutions.com Aurora Computer Services Ltd (Northampton, U.K.) Eye Ticket Corp. (Virginia, U.S.A.) Iridian Technologies (formerlyIriScan, Inc.) Marlton, NJ, U.S.A. and Geneva, Switzerland Saflink (Redmond, WA, U.S.A.)
Retinal scanning (very accurate, very expensive) - Retinal is more intrusive than iris recognition. Eyedentify, Inc. (Delaware, U.S.A.) Microvision, Inc. (WA, U.S.A.) (RSD Retinal Scanning Display) Retinal Technologies, Inc. (MA, U.S.A.)
12
Face Recognition

• Background
• Algorithms
• FERET/FRVT
• Research
• Commercial Products

13
Face Recognition The Basics

• In simplistic form,
• A signature is created from a sensors
  observation
• An algorithm normalizes the signature
• A matcher compares the normalized structure to
  the database.

14
The Algorithms

• Eigenfaces
• Standard Principle Components Analysis (PCA)
• PCA LDA
• LDA Linear Discriminant Analysis
• Combination based on the University of Maryland
  algorithm tested in FERET.
• Baysian
• An Intrapersonal/Extrapersonal Image Distance
  Classifier based on the MIT algorithm tested in
  FERET.
• Elastic Bunch Graphing
• Based on the USC algorithm tested in FERET
• Uses localized landmark features represented by
  Gabor jets

15
Elastic Bunch Graphing
16
FERET and FRVT

• FERET
• DARPA and Army Research Laboratory
• 1994-1996
• A unified means of testing algorithms for easier
  comparison
• FRVT
• Designed by Govt and Law enforcement agencies
• 2000 and 2002
• Tested ability to compare images to those stored
  in a database
• Females and younger people were harder to
  recognize

17
Current Research

• 3-D morphable models
• Not as affected by lighting and pose as is 2-D
• MERL (Mitsubishi) and Ohio State U
• Identical twin Israeli students
• Created a 3-D scanner that uses light to scan the
  image
• Algos measure the distances between points and
  compare to database images
• Factors
• False positives
• Privacy issues
• Environment lighting, movement, etc

18
Commercial Product

• FaceIT (from Visionics / Identix)
• 100
• Developed from an algorithm out of Rockefeller
  University
• Viisage
• From MIT algorithm based on eigenfaces
• TrueFace (from Miros then acquired by Sol
  Universe)
• FaceOK (from Titanium Technology)
• 89
• PC user security

19
Introduction to Fingerprint Recognition

• Fingerprint is the most referred biometric
  mechanism used today.
• Fingerprint has the uniqueness feature the
  studies shows that chance of same fingerprint
  between two individuals (even in twins) is one in
  one billion.
• Fingerprint has been widely adopted (low cost)
  for authentication, identification and criminal
  investigation.

20
Uniqueness of Fingerprint

• Fingerprint is unique because of the two distinct
  feature
• Persistence the basic characteristic of
  fingerprint do not change in time.
• Individuality one over 1 billions !!
• Fingerprints are comprised of various types of
  ridge patterns left loop, right loop, arch,
  whorl and tented arch.
• The discontinuities that interrupt these smooth
  ridge patterns are called Minutia. Minutiae are
  essentially terminations and bifurcations of the
  ridge lines that constitute a fingerprint pattern

21
Fingerprint Capturing and Analysis

• Fingerprint Matching
• Minutae-based
• Correlation based - requires precision location
  of registration point
• Fingerprint Classification
• The technique to assign a fingerprint into one
  of the several pre-specified types already
  established with indexing mechanism
• Fingerprint Enhancement
• It is essential to incorporate a fingerprint
  enhancement algorithm with respect to the quality
  of the fingerprint images in the minutiae
  extraction module in order to ensure the accuracy
  of automatic fingerprint identification/verificati
  on

22
The Identification Workflow of Fingerprint Device
23
DemonstrationIdentix BioTouch 200 USB
Fingerprint Reader

• Hardware - BioTouch 200 USB Fingerprint Reader
• Software - BioLogon for Windows

24
Dynamic Signature Verification

• What is It?
• Uses
• Advantages
• Disadvantages
• Future

25
What is It?

• On-line vs Off-line signature Verification
• Vision-Based, Non-Vision
• Dynamic Signature unique as DNA
• Measures speed, pressure of the pen
• Captures x, y, z location of the writing

26
Uses

• Point of Sale applications
• Workflow automation
• Security
• Authentication replaces password, PIN,
  keycards, identification card
• Financial account opening, withdrawal
• Wireless device security

27
Advantages

• Signatures already accepted as a means of
  identification so people willing to accept
  electronic signature.
• Changes in signing are consistent and have
  recognizable pattern.
• Is not forgotten, lost, or stolen, so simple and
  natural way for enhanced computer security and
  document authorization.
• unique to an individual and almost impossible to
  duplicate.

28
Disadvantages

• Secured authentication
• Difficult to segment strokes as writing styles
  are varied and have no set standard
• Electronic tablets or digitizers are bulky and
  complex.

29
Future

• Administrative Simplification (AS) of the Health
  Insurance Portability and Accountability Act
  (HIPAA)
• IT expenditures
• Frost Sullivan - 5.7M in 2003 up to 123.3M by
  2009
• Mobile phones, Internet, tablet PCs
• PC/Network Access, e-Commerce and telephony,
  physical access and surveillance businesses

30
Thank You

• Have any questions or comments?

Team 3 Steven Golikov Barbara Edington Melanie
Johnson Bashir Amhed Borming Chiang