Illumination Invariant Face Recognition with Whitening Filter and Local Binary Patterns

In this paper we present an approach for the illumination invariant face recognition, which combines 2D spatial whitening filter with the LBP descriptor. Firstly, we preprocess each face image by applying whitening filter in the logarithm domain. Then LBP features are extracted based on the filtered images. After representing image with a LBP operator, the image will be divided into non-overlapped blocks and histograms are used to represent each block. Finally the concatenated histograms are used for recognition. The experimental results on the Yale B and Extended Yale B prove the out performance of the proposed method compared to other existing methods.