Number Of Statistically-based Characteristics And Multiple Classifiers Combination Of Face Recognition Research

Face Recognition Technology is the technology that realizes the recognition of face by computer. Face recognition is a focus of the research in the field of Pattern Recognition and Computer Vision. "This technology involves many related disciplines, and the key technology is the Feature Extraction and Classification Method. This paper develops study based on these emphases, and proposed a Face Recognition Technology which combined the technology based on Wavelet Transform with Multi-feature Multi-classifier. The contents are as follows:The preprocessing phase completed the wavelet decomposition of the original images and the extraction of low frequency images. For the changes of face expression and little mask only hake influence 0n the changes of the high frequency images, low frequency images are still steady under the changes of face expression. So, we can extract face low frequency images by Wavelet Transform. This preprocessing effectively reduced the date quantity of the image, at the same time, Wavelet Transform itself has high computation speed, and it can decrease computational complexity and improve the computational speed under the precondition of guarantee the recognition effect.In the feature extraction phase, we not only extract the feature face feature, singularity point feature and the LAD feature of local self-correlation feature of low frequency images after Wavelet Transform, but also construct mixed feature.In the classification recognition, class each feature by different classifier, and evaluated relevant posterior probability. During the classification, we considered the multi-step classification method, firstly, compute the matching degree of test sample and each training sample, secondly, change them into posterior probability, then, classify by Bayes formula; At last, certify this method by some experiments and ORL face image base, and obtain satisfying recognition effect.