Introduction to Biometrics - PowerPoint - PowerPoint

Document Sample
Introduction to Biometrics - PowerPoint - PowerPoint Powered By Docstoc
					Introduction to Biometrics

             Dr. Bhavani Thuraisingham
           The University of Texas at Dallas

                     Lecture #13
Biometric Technologies: Some Physiological Biometrics

                   October 5, 2005
  Summary of Previous Lectures on Biometrics Technologies
  Some other biometrics
  Course Text Book, Chapter
Summary of Previous Lectures on Biometrics
  Fingerprint Scan
  Face Scan
  Iris Scan
  Voice Scan
Some Other Biometrics Technologies
  Hand Scam
  Retina Scan
  AFIS (Automated Finger print Identification System)
  Multimodal Biometrics
  DNA Biometrics
  Some Behavioral Biometrics
    - Signature recognition
    - Keystroke Dynamics
Hand Scan: Introduction
  This biometric approach uses the geometric form of the hand
   for confirming an individual’s identity.
  Because human hands are not unique, specific features must
   be combined to assure dynamic verification.
  Some hand-scan devices measure just two fingers, others
   measure the entire hand.
  Features include characteristics such as finger curves,
   thickness and length; the height and width of the back of the
   hand; the distances between joints and all bone structure.
  Although the bone structure and joints of a hand are relatively
   constant traits, other influences such as swelling or injury
   can disguise the basic structure of the hand.
Hand Scan: Introduction (Concluded)
  To register in a hand-scan system a hand is placed on a
   reader’s covered flat surface.
  This placement is positioned by five guides or pins that
   correctly situate the hand for the cameras.
  A succession of cameras captures 3-D pictures of the sides
   and back of the hand.
  The hand-scan device can process the 3-D images in 5
   seconds or less and the hand verification usually takes less
   than 1 second.
  Components include: Acquisition hardware, Matching
   software, Storage
Hand Scan: How it works
  Hand geometry scanners such as those made by Recognition
   Systems Inc. take over 90 measurements of the length, width,
   thickness, and surface area of the hand and four fingers--all
   in just 1 second.
  The technology uses a 32,000-pixel CCD digital camera to
   record the hand's three-dimensional shape from silhouetted
   images projected within the scanner.
  The scanner disregards surface details, such as fingerprints,
   lines, scars, and dirt, as well as fingernails, which may grow
   or be cut from day to day.
  When a person uses the scanner, it compares the shape of
   the user's hand to a template recorded during an enrollment
   session. If the template and the hand match, the scanner
   produces an output--it may unlock a door, transmit data to a
   computer, verify identification, or log the person's arrival or
   departure time.
Hand Scan: How it works
  During enrollment, which takes approximately 30 seconds,
   the user places the right hand in the reader three times. The
   unit's internal processor and software convert the hand image
   to a 9-byte mathematical template, which is the average of the
   three readings.
  The user's template may reside in internal memory (capable
   of holding over 27,000 users), or on other media such as a
   hard disk or smart card chip.
  As opposed to such technologies as fingerprint, voice
   recognition, and facial recognition, where a multitude of
   vendors compete via their proprietary technology, hand
   geometry technology is dominated by one company,
   Recognition Systems, Inc. (RSI)
  Finger geometry is led by Biomet Partners.
Hand Scan: How it works (Continued)
  RSI's method for capturing the biometric sample is as
   follows: To enroll, the users places his or her hand palm
   down on the reader's surface.
  The user then aligns his or her hand with the five pegs
   designed to indicate the proper location of the thumb,
   forefinger, and middle finger.
  Three placements are required to enroll on the unit; the
   enrollment template is a representation of the most relevant
   data from the three placements.
  RSI's units use a 32,000-pixel CCD (charged coupled device)
   digital camera, inferring the length, width, thickness, and
   surface area of the hand and fingers from silhouetted images
   projected within the scanner.
Hand Scan: How it works (Concluded)
  Over 90 measurements are taken, and the hand and fingers'
   characteristics are represented as a 9 byte template. source:
   Recognition Systems, Inc.
  Biomet Partners' technology is similar, but draws on the
   shape and characteristics of the index and middle finger. The
   data is saved as a 20 byte template.
  Hand geometry does not perform 1-to-many identification, as
   similarities between hands are not uncommon.
  Where hand geometry does have an advantage is in its FTE
   (failure to enroll) rates, which measure the likelihood that a
   user is incapable of enrolling in the system. Fingerprint, by
   comparison, is prone to FTE's due to poor quality
   fingerprints; facial recognition requires consistent lighting to
   properly enroll a user.
Hand Scan: Template Generation and Matching
  Distinctive features include height, width, thickness of the
  Distinctive features of the hand and finger are extracted from
   a series of 3-D images and recorded into a small templates
  False matching and false non-matching are possible due to
   the fact that hands may swell and undergo changes
Hand Scan: Applications
  Hand geometry is currently among the most widely used
   biometric technologies, most suitable for access control and
   time and attendance applications.
  Hand scan is used reliably at thousands of places of
   employment, universities, apartment buildings, and airports -
   anyplace requiring reasonably accurate, non-intrusive
  The nature of hand geometry technology is such that most
   projects are fairly small-scale and involve only a handful of
   readers, but there are some projects which incorporate
   dozens of readers.
Hand Scan: Deployments
  INSPASS (Immigration and Naturalization Service Passenger
   Accelerated Service System) project, one which allows
   frequent travelers to circumvent long immigration lines at
   international airports.
  Qualified passengers, after enrolling in the service, receive a
   magstripe card encoded with their hand scan information.
   Instead of being processed by passport control personnel,
   INSPASS travelers swipe their card, place their hand, and
   proceed with their I-94 to the customs gate.
  Nearly 50,000 people have enrolled in the service, and
   approximately 20,000 verifications take place every month.
   Travelers from 30 different countries are qualified to register
   for INSPASS; pending budgetary constraints, the near-term
   objective is to rollout the INSPASS project to over 20 airports
   in the U.S.
Hand Scan: Market Size
  Hand geometry is projected to be one of the slowed growing
   biometric technology through 2007.
  Because the range of applications in which hand geometry is
   typically limited to access control and time and attendance, it
   will draw a progressively smaller percentage of biometric
  Overall, hand geometry revenues are projected to grow from
   $27.7m in 2002 to $97.4m in 2007. Hand geometry revenues
   are expected to comprise approximately 2.5% of the entire
   biometric market.
Hand Scan: Strengths and Weakness
  Strengths
    - Ease of use.
    - Resistant to fraud .
    - Template size - Using RSI, a template size of 9 bytes is
      extremely small
    - User perceptions – non-intrusive
  Weaknesses
    - Static design - largely unchanged for years.
    - Cost
    - Injuries to hands
    - Accuracy, hand geometry, in its current incarnation,
      cannot perform 1-to-many searches, but instead is limited
      to 1-to-1 verification.
Retina Scan: Overview
  Completely different from Iris Scan
  Camera captures the image of the retina
  Movements affects the images
  Need about 3 – 5 images for enrollment
  Distinctive features include network of blood vessels
  Glaucoma and other conditions may affect retina scan
  Template generation process will map the unique network of
   blood vessels into a template
  Template is about 96 bytes
  Usually does one-many identification
  Good for highly secure environments
Retina Scan: Overview (Concluded)
  Strengths
    - Resistance to false matching
    - Stable characteristics
  Weakness
    - Difficult to use
    - User discomfort
    - Limited applications
Retina Scan: Details
  Retinal scanning analyses the layer of blood vessels at the
   back of the eye.
  Scanning involves using a low-intensity light source and an
   optical coupler and can read the patterns at a great level of
  The user looks through a small opening in the device at a
   small green light. The user must keep their head still and eye
   focused on the light for several seconds during which time
   the device will verify his identity. This process takes about 10
   to 15 seconds total.
  There is no known way to replicate a retina, and a retina from
   a dead person would deteriorate too fast to be useful, so no
   extra precautions have been taken with retinal scans to be
   sure the user is a living human being.
Retina Scan: Details (Continued)
  Retina scan is actually one of the oldest biometrics as 1930's
   research suggested that the patterns of blood vessels on the
   back of the human eye were unique to each individual.
  While technology has taken more time than the theory to be
   usable, EyeDentify, founded in 1976, developed The
   Eyedentification 7.5 personal identification unit, the first
   retina scan device made for commercial use, in 1984.
  At this time, they are still the primary company for retinal
   scan devices
  Retina scan is used almost exclusively in high-end security
  It is used for controlling access to areas or rooms in military
   installations, power plants, and the like that are considered
   high risk security areas.
Retina Scan: Details (Concluded)
  Retina scan devices are provide accurate biometric
  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 the 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.
  Automated Fingerprint Identification System (AFIS)
   technology is used in a variety of law enforcement and civil
  In law enforcement, fingerprints are collected from arrested
   subjects and searched against local, state, regional, and/or
   national fingerprint databases.
  The subject's ten fingerprints are acquired either through the
   traditional ink-and-roll method or through an optical livescan
   system, consisting of a sizeable fingerprint scanner, PC, and
   imaging and transmission software
  The electronic fingerprints are submitted, along with
   demographic data, to identify or verify the identity of the
AFIS (Concluded)
  Searches may take minutes, hours, or days, depending on the
   quality of the information submitted, the size of the database
   being searched, and the entity requesting the search.
  Law enforcement searches often return candidate lists used
   to determine which of several possible matches is the best
  Most widely used biometric technology
  AFIS is different from fingerprinting systems
     - AFIS captures (in addition to templates) and uses image
       analysis algorithm
Multimodal Biometrics
  A multimodal biometric system uses multiple applications to
   capture different types of biometrics.
  This allows the integration of two or more types of biometric
   recognition and verification systems in order to meet
   stringent performance requirements.
  A multimodal system could be, for instance, a combination of
   fingerprint verification, face recognition, voice verification
   and smart-card or any other combination of biometrics.
  This enhanced structure takes advantage of the proficiency of
   each individual biometric and can be used to overcome some
   of the limitations of a single biometric.
DNA Biometrics
  Proving that a suspect's DNA matches a sample left at the scene of a
   crime requires two things: Creating a DNA profile using basic
   molecular biology protocols; Crunching numbers and applying the
   principles of population genetics to prove a match mathematically
  Humans have 23 pairs of chromosomes containing the DNA
   blueprint that encodes all the materials needed to make up your
   body as well as the instructions for how to run it. One member of
   each chromosomal pair comes from your mother, and the other is
   contributed by your father.
  Every cell in your body contains a copy of this DNA; While the
   majority of DNA doesn't differ from human to human, some 3 million
   base pairs of DNA (about 0.10 percent of your entire genome) vary
   from person to person.
  The key to DNA evidence lies in comparing the DNA left at the scene
   of a crime with a suspect's DNA in these chromosomal regions that
   do differ.