Feature Extraction And Reasearch In The Application Of Face Recognition

Face recognition is an extremely challenging issue. The research of Face recognition has not only the theoretical research value but also wide application. The theory of feature extraction and its application in the field of face recognition are the main content of this thesis. The face recognition algorithms based on the global feature extraction and local feature extraction are the research emphasis. In addition, some improved feature extraction algorithms are proposed. The main content of this paper is as follows:First, two face recognition algorithms based on global feature extraction (eignface algorithm and Fisherface algorithm) are introduced.The two algorithms are implemented in ORL database and as a result the recognition rate of Fisherface algorithm is higher than eignface algorithm.Second, generalized principle component analysis is introduced. an improved method is also given by doing two generalized principle component analysis in the horizontal direction and the vertical direction. In this way, we get two face characteristic matrixes. Then, according to some perfect weight, we can together the two characteristic matrixes and make good use of all the face imformation. The experiment in ORL database shows that the new algorithm improves the recognition rate.Third, the local feature extraction theory is introduced. The experiment in ORL database shows that the local feature extraction algorithm can use the local features of face fully to improve the recognition rate. It is also more stable in changing light conditions.Fourth, considering the feature extraction problem as a combinational optimization problem, new feature extraction algorithm based on the genetic algorithm and binary partical swarm optimization algorithm are proposed. The experiment shows that the new feature extraction algorithms are able to compress the face characteristic space and get higher recognition rate.