Illumination-Robust Face Recognition based on Gabor Feature Face Intrinsic Identity PCA Model

Abstract

Robust face recognition under various illumination environments is essential for successful commercialization. Feature-based face recognition relies on a good choice of feature vectors. However, there is no feature vector invariant under illumination changes even though some feature vector such as Gabor feature vector is relatively robust to variations of illumination. Also, illumination normalization techniques cannot eliminate illumination effects completely. In this paper, we propose an illumination-robust face recognition method based on the face Gabor intrinsic identity PCA model. We first analyze face Gabor feature vector space and construct a face Gabor intrinsic identity PCA model which is independent Of illumination effects and propose a face recognition method based on it. Through experiments, it is shown that the proposed face recognition based on face Gabor intrinsic identity PCA model performs more reliably under various illuminations and pose environments.