Feature Extraction And Face Recognition Techniques Based On Subspace Analysis

Face recognition is a hot area in pattern recognition and image process recently, and the research for it will be benefit to the progress of pattern recognition and information security. Feature extraction is the elementary problem in the area of pattern recognition, and it is the key to solve the problem of image identification. Linear projection analysis, including principal component analysis (or K-L transform) and fisher linear discriminant analysis, is the classical and popular technique for feature extraction. The appearance of kernel trick which can derive non-linear feature improves the develepment of face recognition, and feature extraction technology such as kernel principal component analysis and kernel fisher linear discriminant analysis get more attention. In addition, principal component analysis (or K-L transform) and fisher linear discriminant analysis need much computering time, but appearance of the 2 dimensional principal component analysis and 2 dimensional fisher linear discriminant analysis could solve the problem. All the methods mentioned above is based on subspace, so they are called subspace methods. In this paper, focusing on subspace analysis methods, some new algorithms are developed. And, these algorithms are verified to be effective in the application of face identification.Firstly, A new face recognition method based on SOPCA (second-order PCA) and KPCA is proposed in this paper. K-L transform method is used to transform initial images, then the second-order face image is got through rebuilding images, and KPCA is used to get two kinds of feature vectors for the initial image and its second-order face image respectively. Lastly, the two kinds of vectors of everyone are combined into a longer vector .To verify the efficient of the method, experiment is tested on ORL face database and experiment result shows that this face recognition method is more...