Fusing DCT and LBP features for face recognition

A novel method by fusing discrete cosine transform (DCT) and local binary pattern (LBP) features is proposed for face recognition in this research. The primary information of the face image was centralized in a small number of DCT coefficients, which were used as the frequency feature of the face. The face was divided regularly into small regions, from which LBP code histograms were computed and concatenated into a spatial global histogram used as descriptor vector of the face. The descriptor vector was dimensionally reduced by PCA. Then, the DCT features and the LBP features were fused posterior to the normalization. The experiments on ORL face database show the improvability of the proposed scheme on the methods using just single DCT or LBP feature for face recognition.