Face Recognition Based On Optimized Bayesian Method

Face recognition technology is the cross discipline of applied mathematics, digital image processing and pattern recognition and is emerging as an active research area in the field of biological pattern recognition. The biological characteristics are the inherent attributes of human beings, which have strong self-stability and individual independency. They are the ideal source of information for ID verification.The process of face recognition can be divided into four steps:face detection, preprocessing, feature extraction, classification. In this paper we introduce the two stages of preprocessing and feature extraction in great detail. Feature extraction is the main problem of face recognition, the feature should ensure itself has representative attribute, more information, less redundancy and some steady attribute resist interference. Compared with other modern recognition, face recognition is the most natural and direct identity authentication method, so it has direct, friendly and convenient characters. First, we introduce the method of feature extraction, for example Principle Component Analysis, Linear Discriminate Analysis, Singular Value Decomposition and the combination of these methods and so on.Later, we compared the performance of these methods through experiments.In the recognition stage we mainly introduce the face recognition based on Bayesian. Because the modern vectors are very high-dimensional, so we should use the method of PCA and LDA to reduce the dimensional. When select the intrapersonal subspace, we add the adaptive factors to the average intrapersonal subspace obtain the specific intrapersonal subspace. So the recognition rate is improved.Later we introduce the maximum entropy covariance selection method.In its application in Bayesian method, it not only solve the limited training samples but also improve the recognition rate.