In this paper,the key technologies of face recognition system are introduced,and the research of face recognition technology under different illumination is reviewed;the current illumination preprocessing methods include wavelet-based processing,Retinex method,etc.,and the performance differences of the same illumination preprocessing method in different face test libraries are compared and analyzed;the specific preprocessing method can achieve the ideal recognition effect for the specific attribute database.An adaptive adjustment algorithm for face illumination is proposed,and a fuzzy recognition model is established for algorithm experiments to improve the accuracy of face recognition under non-ideal conditions,so as to select clear face images to improve the accuracy of face recognition in moving scenes.The accuracy of face recognition reaches 97.21%by using the design algorithm,which is worth popularizing.
维普
万方数据
