Face recognition based on completed local binary pattern
基于完整LBP特征的人脸识别*

This paper presented a method of face recognition based on completed local binary pattern. CLBP operators was decomposed into three operators, namely CLBP_C,CLBP_S and CLBP_M, are proposed to code the C,S and M features, respectively. First,it used CLBP operator to extract the CLBP histogram from block grey-level face images. Then combined the CLBP_C,CLBP_S and CLBP_M codes to form the CLBP feature map of the original image and evaluated the goodness between tow CLBP histograms. Finally, face recognition based on the nearest neighbor principle. The simulation experiments illustrate that this method has better recognition rate and more robust than LBP on the ORL and YALE face database.