BIOMETRICS

1
BIOMETRICS
Maggie Perry
2
What is Biometrics?

• The term biometric can refer to two things
• A characteristic a measurable biological and
  behavioral characteristic that can be used for
  automated recognition
• A process an automated method of recognizing an
  individual based on measurable biological and
  behavioral characteristics

3
(No Transcript)
4
Fingerprint

• Fingerprints are unique and consistent.
• Fingerprints consist of a pattern of peaks and
  valleys.
• Various kinds of sensors
• can be used.
• 2 categories of matching
• software- pattern and
• minutiae-based.

5
Hand Geometry

• The hand is placed on a plate with 5 guiding
  pegs.
• 3 images are taken and over 900 measurements are
  made.
• The measurements taken are compared to those
  stored.
• This is only used
• for verification.

6
Iris

• Iris patterns are not genetically linked and are
  unique.
• Infrared digital camera used to capture image.
• The image is converted into a IrisCode.
• In recognition the
• IrisCodes on file and
• the new code are
• compared.

7
DNA

• DNA is not currently used as a biometric for
  several reasons
• Analysis and comparisons cannot be done in real
  time.
• Actual samples must be taken, not just images.
• DNA is the biometric of the future.

8
Face

• There are 2 different approaches- geometric and
  photometric.
• 3 algorithms are commonly used.
• Principal Components
• Analysis (PCA)
• Linear Discriminant Analysis (LDA)
• Elastic Bunch Graph
• Matching (EBGM)

9
Voice

• Voice recognition is both a physiological and
  behavioral biometric.
• 2 categories of voice recognition.
• Text Dependent
• Text Independent
• Performance is
• inconsistent but
• changes are still
• being made.

10
Dynamic Signature

• This is very different from electronic signature
  capturing.
• Characteristics vary from
• program to program.
• Some programs are
• adaptable and can learn.
• Very unique and difficult to
• trick unlike graphical images.

11
Comparing Biometrics

• Uniqueness
• Permanence
• Collectability
• Performance
• Acceptability
• Circumvention

12
(No Transcript)
13
Biometric Systems

• Composed of 5 parts
• Sensor
• Signal processing algorithm
• Data storage
• Matching algorithm
• Decision process

14
Applications

• Security
• Identification
• Disney World!!!

15
Works Cited

• http//en.wikipedia.org/wiki/ImageBiometrics_trai
  ts_classification.png
• http//www.biometrics.gov/docs/irisrec.pdf
• http//www.biometrics.gov/docs/handgeometry.pdf
• http//www.biometrics.gov/docs/fingerprintrec.pdf
• http//www.biometricgroup.com/reports/public/repor
  ts/dna.html
• http//www.biometrics.gov/docs/speakerrec.pdf
• http//www.biometrics.gov/docs/facerec.pdf
• http//www.biometrics.gov/docs/dynamicsig.pdf
• http//www.biometrics.gov/docs/biooverview.pdf