This document discusses biometrics and biometrics research at Clemson University. It defines biometrics as automated methods of recognizing individuals using biological and behavioral characteristics. It describes several types of biometric recognition including fingerprint, iris, facial, voice, hand geometry, signature, and retina recognition. It also discusses commercial, government, and forensic applications of biometrics and constraints on biometric systems like lighting and pose. Finally, it outlines biometrics research at Clemson University focused on usable and unconstrained biometrics including aging effects, demographics, periocular recognition, and ear recognition.

2. Overview
 What is Biometrics?
 Types of Biometric Recognition.
 Applications of Biometric Systems.
 Types of Authentication.
 Constraints on Biometrics.
 Biometric Research at Clemson University.

3. What is Biometrics?
 Automated method for recognizing individuals based
on measurable biological and behavioral
characteristics

4. Finger Print Recognition
 Minutiae
 Pattern Matching
 Problems: sometimes
unusable

5. Vascular Pattern Matching
 LED infrared light
 Fingers and back of hand
 Not completely viable

6. Iris Recognition
 Uses infrared light
 Converts Images to vectors
 Needs further
development

7. Facial Recognition
 Location and position of
facial features
 Dependent on background
and lighting conditions

8. Voice Verification
 Factors: pitch, intensity,
quality and duration
 Text dependent
 Text independent
 Problems: include
background noise

9. Hand Geometry
 Scan both sides of hand
 Primarily used for
verification
 Not as accurate as other
methods

10. Dynamic Signature
 Factors: velocity,
acceleration and speed
 Mainly used for verification
 Problems: forgers could
reproduce

11. Retina Recognition
 One of the most secure
means of biometrics
 Unique to each person
 Unique to each eye
 Problems: require effort on
the part of subjects

12. Other Types
 Keystroke
 DNA

13. Commercial Applications
 Computer login
 Electronic Payment
 ATMs
 Record Protection

14. Government Applications
 Passport control
 Access Control

15. Forensic Applications
 Missing Persons
 Corpse identification
 Criminal investigations

16. Type of Authentication
 Authentication
 1:1
 Verification
 1:N

17. Constraints on Biometrics
 Typical “Constrained”
Image
 Constraints:
 Lighting
 Distance
 Pose
 Expression
 Time Lapse
 Occlusion

18. Biometrics Research at
Clemson University.
 Biometric and Pattern
Recognition Lab
 Goals:
1. Usable Biometrics
2. Unconstrained Biometrics

19. Biometrics Research at
Clemson
 Aging Research

20. Biometrics Research at
Clemson
 Demographics
 Older vs. Younger
 Males vs. Females
 Geographic origin of
algorithms

21. Biometrics Research at
Clemson
 Periocular Region
Recognition
 Texture, color, eye shape
 Overcome facial occlusion

22. Biometrics Research at
Clemson
 Ear Recognition
 Not affected by aging or
expression
 Covert collection of
images
 Little research performed