Face recognition

Abstract

Face recognition played an important part in daily lifestyle, security systems (Criminal Identification), and biometric purposes etc. The high dependency of face recognition technology in today's society has hence aroused much interest in researchers to develop a reliable face recognition system which leads to the objective of the project, to study the Principle Component Analysis (PCA) and to analysis the different variable such as threshold, number of eigenvectors. The author will also describe in detail description on every part of the process including the mathematical method and algorithm of PCA. The face database that will be used is AT&T face database. The literature review will cover other techniques and comparision of the accuracy between PCA, ICA and LDA. Vigorous testing will be done on the AT&T face database to analyse the relationship between threshold, number of eignvectors, number of training and testing images towards the recognition rate of the system mainly using PCA. The testing includes comparison of results between the Eucladian and Mahalanobis algorithms. From the results data, in general, the Mahalanobis Distance give better results as compared to the Eucladian Distance.