A New Method of Face Recognition

During face recognition,there exists a serious problem that it is very difficult to distinguish faces of various gestures.The discrete cosine transform(DCT) is introduced to reduce negative effects.K-L transformation can be used to compress images.According to primary component analysis(PCA),we utilize K-L to decrease data dimensions,i.e,only parts of vectors are used for data representation.Finally,RBF neural network is enrolled to recognize various faces.The task to train RBF is finished by PSO.It is proved by our experience that this method has a high accuracy.