FACE RECOGNITION USING PCA, GABOR FILTER AND SVM TECHNIQUES

Face recognition is the hot research topic from last few years but still it has become a difficult problem. The main challenges faced by the researchers are variation caused due to different expression and poses. The main feature that can be used to extract the features from variant images that are caused because of different variations is Gabor Wavelets. This technique makes it possible to use their facial image of person to authenticate him into a secure system. In this paper the Principle Component Analysis face recognition algorithm is used for recognizing the faces this technique effectively and efficiently represents pictures of faces into its eigen faces components and these eigen face components form eigen faces these eigen faces are the ghost images of original images. The significant feature known as eigen faces don’t necessarily correspond to features such as eyes, ears and noses. I t provides the ability to learn and later recognizes new faces in an unsupervised manner. The efficiency and robustness of a proposed algorithm is extensively tested using Standard Database (UCI), Non - Face and Own databases. An automatic user identificati on system consisting of detection, recognition and user management modules have been developed. The feature vector based on gabor filter are used as the input of the face/Non - face classifier, which is SVM on a reduced feature subspace extracted by using PC A.