Face Recognition Based On Global And Local Feature

Face recognition technology because it involves more extensive areas of research and application scenarios is currently become a hot research topic in the research or whether it has an important value in the commercial. Experienced decades of development there are a large number of researchers separately for different practical applications put forward their own face recognition algorithm. Based on the research base extraction algorithms were based on global and local characteristics, the proposed integration of global and local features of face recognition algorithm, and the algorithm with whole or partial recognition rate algorithms were compared. The main contents of this paper are as follows:(1) The main face recognition algorithms are purely based on global feature extraction algorithm or based on local feature extraction algorithm and some algorithms improved, many researchers put forward the future trend of face recognition is a fusion of a variety of algorithms.In this paper, the method of face recognition based on the integration of the wh ole feature and the local feature is used to improve the face recognition rate. First, PCA algorithm is used to extract face image as a whole an important feature information. Then the local binary pattern algorithm to obtain some important information and then use Bayesian fusion strategy for effective integration.When comparing the experimental results and the results take the same global features extraction algorithm and the same local feature extraction algorithm of feature level fusion algorithms were compared, such result is more reliable. The contrast results prove that the recognition rate of this paper is higher than the recognition rate of principal component analysis under the same training sample number. The highest recognition rate of this method is higher than that of local binary pattern. Of course, the recognition rate of this algorithm is higher than the simple feature level fusion of the same feature extraction algorithm.In a word, Fusing the two effective information can combine the advantages of the two to improve the rate of face recognition purposes.(2) In order to train, test, identify and select the image of the face database,this paper designe a interface of a face recognition system. And the left parameter of the interface always shows the data of the last experiment, and can be conveniently recorded for the experiment data, so that we can choose better parameters in experiment.