(Initial report by Nont, 11/22/05)

Biometric ["bios" + "metron" ("life" + "measure")]

Overview:

"Biometric technologies are defined as "automated methods of identifying or authenticating the identity of a living person based on a physical or behavioral characteristic." Unique physical traits, such as fingerprints, iris scans, voiceprints, faces, signatures or the geometry of the hand, can be used. All of those technologies share a methodology involving enrollment and verification. At enrollment, the person offers a "live sample" of the biometric, such as a fingerprint. This is scanned electronically, processed and stored as a template. To confirm the identity at a future time, the individual presents the live sample, which is matched against the stored template. In a "1:many" search, the individual presents only the live sample, and the database is searched for a match. This is called identification. In a "1:1" search, the user presents a name or personal identification number along with the live sample. The system checks the live sample only against samples stored under that name or PIN. This is called verification or authentication."

(Source: DOD Biometrics Management Office)

Type of biometrics:

Physiological and Behavioral

PhysiologicalBehavioral
IrisSignature
FingerprintKeystroke
Hand (including knuckle, palm, vascular)Voice
FaceGait
VoiceMouse Dynamics
Retinae.g.
DNA Pattern 
Odor, Earlobe, Sweat pore, Lips 
Skin Spectrum 
e.g. 

Biometrics can moreover be defined as either Active, or Passive.

Active biometrics: (require person cooperation and will not work if they deny their participation in the process)

  • Fingerprint
  • Hand geometry
  • Retina scanning
  • Signature recognition

Passive biometrics: (do not require user's participation and can be successful without a person even knowing that they have been analyzed)

  • Voice recognition
  • Iris recognition
  • Facial recognition

Verification or Identification:

Basically, a biometric system is a pattern recognition system which makes a personal identification by determining the authenticity of a specific physiological or behavioral characteristic possessed by the user. An important issue in designing a practical system is to determine how an individual is identified. Depending on the context, a biometric system can be either a verification system or an identification system.

Traditional verificationIdentification
BinaryPattern Recognition
Password, Pin, BiometricBiometric
Yes/noScore (degree of correlation between the match template and the reference template)
1:11:Many
"Are you whom you claim to be?""Who are you?"

Working Process:

Enrollment --> Submission --> Acquisition device --> Biometric sample --> Feature extraction --> Template --> Matching --> Score --> Threshold --> Decision

TechnologyBiometric SampleFeature Extracted
FingerprintFingerprint imageLocation and direction of ridge endings and bifurcations on fingerprint
Voice recognitionVoice recordingFrequency, cadence and duration of vocal pattern
Facial recognitionFacial ImageRelative position and shape of nose, position of cheekbones
Iris-scanIris ImageFurrows and striations in iris
Retina-scanRetina ImageBlood vessel patterns on retina
Hand geometry3-D image of top and sides of hand and fingers Height and width of bones and joints in hands and fingersHeight and width of bones and joints in hands and fingers
Signature-scanImage of signature and record of related dynamics measurementsSpeed, stroke order, pressure, and appearance of signature
Keystroke-scanRecording of characters typed and record of related dynamics measurementsKeyed sequence, duration between characters

Applications (in general):

  • Time Attendance
  • Access control
  • Identification card
  • Immigration checks
  • Police records
  • Customer identification, Loyalty programs
  • Security systems
  • Patient management in hospitals
  • Public welfare programs management
  • Preventing identity theft
  • Membership management in clubs, libraries etc.
  • Enterprise network security
  • Transaction management
  • Banking, ATMs

Example Applications

Siemens ID Mouse

I couldn't use this mouse with my PC. I guess that it's because I use window XP, but the driver that I have (version 4) is for window98, window 2000, window NT only.

The fingerprint sensor which is used for this mouse is Infineon FingerTIP.

Demonstration Application Note:

Idea: To experiment using an infrared camera as a sensor for a simple shape recognition system to identify each different user in my apartment

screenshot from my prototype

For the demonstration application for this report, I tried to experiment the simple prototype of the simple indentification system based on shape recognition by using infrared web camera with Max/Jitter.

Hardware:

  • Infrared Web Camera
  • Computer
  • Incandescent Lamp
  • Controlled Environment Space

Software

  • Max/Jitter + cv.jit

The key of this system is “cv.jit.undergrad” which provides the simple pattern recognition algorithm object for using in max environment. (This is a realy great computer vision object.)

The infrared camera that I modified from a logitech web camera.

Using incandescent lamp to set the lighting in the environment (I need IR light from incandescent lamp for this prototype.)

User testing

(More details soon)

Future works:

The reason that I chose to report about biometrics is because I would like to study the basic idea behind this system before using it in my independent study project, which is about the scenario planning for the future of personal space. I plan to apply biometric system to the architectural and spatial design in my next projects as well (about the space that has memories.)

Note:

From this study, I found that,generally, biometric is in a subset of the "sensor + pattern recognition". Most of biometrics require using pattern recognition system with the sensor to identify the individual. However, the potential of pattern recognition is not just for the biometrics or the security applications, there are many kinds of products which use this pattern recognition system as well (e.g. robot, performance/sport practice instrument, or game controller.) Personally, I am really interested in using simple pattern recognition with sensor to develop my future projects. Personally, this is a new world for me, and I wish I will be able to understand more about pattern recognition in near future (I wish.)

More Resources:

High-level overview of biometric technologies(by National Defense University, Fort McNair, Washington, DC. )

The Biometrics Resource Center, National Institute of Standards and Technology

How fingerprint scanner works (www.howstuffworks.com)

Rosystem

Standard Deviation(mathworld)

Standard Deviation(wikipedia)

Acknowledgements:

I would like to thank Mr. Jean-Marc Pelletier, IAMAS. for the very helpful information about cv.jit.undergrad, and Tom Igoe for lending me a Siemens ID mouse.