| Abstract: | Face recognition, one of the biological recognitions, has received extensive concern due to its secrecy and friendly cooperation. Gabor wavelet is an important tool in face feature description. In order to reduce the loss of useful information during down sampling, this work puts forward a Gabor feature representation method based on block statistics, which enhances the efficiency of Gabor feature representation. This study was designed to explore face recognition algorithms on the basis of highly recognizable and real-time collaborative representation. Experimental results indicated that, the face recognition based on block Gabor feature collaborative representation not only guaranteed the calculation speed, but also took full advantage of the robustness of Gabor feature. Besides, the block Gabor feature containing more details further improved the recognition rate. |
