An architecture for face classification

Biometric classification techniques, in which measurements of the physical characteristics of some distinguishing feature of the individual are made, have increasing application in the security field. Examples of biometric techniques include systems based upon signatures, fingerprinting, retinal scanning, hand geometry, and facial feature extraction. Of crucial importance in the application of such techniques is their acceptability to those who will be subjected to them, with obvious advantages for techniques that do not involve the physical constraint of the subject, and are otherwise unobtrusive. Automatic face identification offers the advantages discussed above. The authors discuss previous research into face identification, and present a face identification algorithm that would be suitable for implementation in a workstation security environment