MPhil Thesis Defence


"Personal Identification/Authentication by using Hand Geometry"

By

Mr. Chin-Man Wong


Abstract

This thesis presents an approach for personal
identification/authentication with the use of Hand Geometry. There have
been some related works in this area, but most of them require pegs when
capturing hand image and only the upper half of the hand area is used for
encoding feature vector. We present a new feature extraction method that
utilize the whole hand and can handle variation of hand placement. No pegs
are needed to fix the hand position. The captured image will be processed
to form a binary (black and white) and lines images. A Feature vector with
29 features will be extracted from these two images. Further feature
selection process reduces the size of feature vector to 17. Noisy samples
can be eliminated by the outlier removal process. We present promising
results, which show up to a 98.7% success classification rate and 2.4%
Equal Error Rate (EER) on a 100 individuals hand image database.


Date:				Thursday, 22 May 2003

Time:				10:00a.m.-12:00noon

Venue:				Room 1505
				Lifts 25-26

Committee Members:			Dr. Helen Shen (Supervisor)
				Dr. Dit-Yan Yeung (Chairman)
				Dr. Brian Mak


**** ALL are Welcome ****