Abstract. In order to the practical face recognition system using in an uncontrolled environment exist many outside interference factors. In this paper, it is proposed that a novel facial standardization method to extract the core region of face samples, furthermore, simultaneously adjust the facial posture and achieve face alignment. It can avoid "mis-alignment disaster" during the feature matching process. In this paper, a video face recognition system of small samples capacity has been established to test this method. The experiments show that the method can make the face samples optimal in the uncontrolled environment and enhance the recognition rate of the face recognition system.

No abstract