Design And Implementation Of Face Recognition System Based On EHMM

Face recognition technology is an important research topic in the fields of pattern recognition and artificial intelligence.it also has a broad prospect application in information security,criminal detection,identity authentication.After years of development,face recognition technology has made considerable progress and has a high rate of recognition in the ideal case.However,because of the influence of illumination change,expression change and low resolution in actual environment,the face recognition rate dropped significantly.particularly, the effect of illumination change is the most serious.Therefore,there are lots of problems need to solve if we want to use the face recognition technology widely in commercial and family life.In order to solve the problem of illumination change in face recognition.lt first use the Retinex to reduce the impact of illumination change in the part of image preprocessing,then use the local binary pattern (LBP) and discrete cosine transform(DCT) to extract face feature in the feature extraction partthrough use LBP and DCT extract the local texture feature which is not sensitive to the illumination change and the global feature respectively to get the complementary feature of global feature and local feature,finally put the feature vector into the embedded hidden markov model to practice and recognition.Lots of experiments have been done on ORL.Yale B and homemade face database,in order to verify the performance of the algorithm from this paper.Results show that the algorithm has high recognition rate and strong robustness to the illumination change.In the end of the paper,a detailed description of each sub-module program design of the process of face recognition combined with face recognition algorithms proposed,related to use of software engineering methods,and finally achieves an face recognition system based on EHMM.