An Efficient Recognition Approach to Human Faces
一种有效的人脸识别方法

Human face recognition is a very difficult and important problem in the field of pattern recognition. In this paper, the characteristic images and projective images are first introduced. The projective coordinates of projective image on the characteristic images are presented to describe the intrinsic attributes of human faces, Then, a feature extraction method of human faces based on the projective images is proposed. Finally, a hierarchical distance classifier is designed to recognize human faces. Experimental results show that the feature vector based on the projective images has good stability, discriminant ability and recognition rate.