Illumination invariant face recognition under various facial expressions and occlusions

Abstract

This paper presents a new approach that increases face recognition performance using various facial expressions in the presence of illumination variations and occlusions. The new approaches use PCA and LDA with the combination of the preprocessing techniques of histogram equalization and mean-and-variance normalization in order to nullify the effect of illumination changes and any occlusions present which are known to significantly degrade recognition performance. To be consistent with the research of others, our work has been tested on the JAFFE database and its performance has been compared with traditional PCA and LDA.