Evaluation of CIE-XYZ system for face recognition using kernel-PCA

Abstract

paper evaluates the performance of face recognition with different CIE color spaces. The XYZ and L*a;*b* color spaces are compared with the gray image (luminance information Y). The face recognition system consists of a feature extraction step and a classification step. The Kernel-PCA is used to construct the feature space. Kernel-PCA is a nonlinear form of Principal Component Analysis (PCA). The k-nearest neighbor classifier with cosine measure is used in the classification step. Experiments using FEI color database with 200 subjects, show that the b* color component can improve the recognition rate.