摘要：In order to achieve more local facial feature, a kind of sub image face recognition method based on RS-Sp CCA feature information fusion is proposed in this paper. According to take samples for the local facial feature of sub image and use CCA to fuse the global facial feature and the local facial feature information after sampling, the global feature of image can be fully used to construct much more different kinds of component classifiers. Then, make experimental analysis on databases of 3 standard facial data sets. At last, the results show that sub image face recognition method based on RS-Sp CCA feature information fusion is better than the simple feature sampling method and feature fusion method. Besides, it can efficiently improve the face recognition rate.
