Research And Application Of Face Detection Based On Lighting Per-processing

With the development of economic and information globalization, Video monitoring hasbecome a routine measure of preventive security in the current society, But in the process ofcapturing images, Light isn’t artificial control, casing images’ poor contrast and lowresolution, which brings considerable difficulties to face detection, face tracking and facerecognition. This paper study the face detection and the face recognition algorithm based onthe light preprocessing algorithm.Firstly, this paper describes several common light pre-processing algorithms, analysisand compare with the advantages and weaknesses of these algorithms. the phenomenon suchas white, color distortion, low contrast and a long time, which might be produced while usingretinex algorithm to process face images, this paper proposes a new and multi-scale retinexwith color restoration algorithm. On the basis of studying of MSRCR theory, we use the fastFourier algorithm into the original algorithm to make the speed faster than that of thetraditional method. The experimental results show that this method can better restore theinformation of dark area on images, which further improves the face recognition rate.Secondly, after lighting pre-processing, the face unique color characteristics are used toachieve the fast face detection. Skin model is established on the color space of improvedYcbcr; skin detection and threshold segmentation are executed according to improved skinmodel; In order to more accurately, this paper presents an optimization algorithm based on thecandidate face region segmentation. The algorithm is analysis and process of mathematicalmorphology of the candidate face region segmentation. The face area and pre-production ofthe face template are matched; at last, the matched area is marked by rectangle.Finally, in this paper, we use PCA to solve face recognition, and Visual Studio2010,Matlab mixed programming has developed a very practical, convenient operation of facedetection and recognition system based on lighting pre-processing.