Research And Realization On Face Recognition

The automatic human face recognition is a technology that using computer to analyze the human face images and extracting effective features from the human face images, then to recognize them. Compared with other biometric methods, such as fingerprint, palm, iris, and DNA, human face recognition is more friendly, direct, and imitate to humane. Therefore, in recent years, face recognition is getting more and more attentionI did some researches on both face features extraction and features recognition. The total work of the paper research includes:Firstly, researches on typical statistical features extraction methods, including PCA, ICA, LDA, DCT, then developed a feature combination of ICA, LDA and DCT.Secondly, researches on SVM and its application on face recognition. Development of a kernel function based on cosh function, and its application on face recognition. Development of a characters selection method of SVM based on Discrete Particle Swarm Optimizer, which effectively accelerates the characters selection.Thirdly, establishment of the real time face recognition system, using ICA for feature extraction and SVM for classification. Validation of the feasibility and activities of the related arithmetic.