中文摘要：

为了消除光照变化对人脸识别的影响，提出了一种新的基于小波的光照归一化算法。首先对人脸图像进行三级小波分解，获取低频和高频系数；接着对低频成分直方图均衡化，减弱光照的影响，同时对高频成分阈值

英文摘要：

In order to eliminate the effect of illumination variation on face recogntion,an improved wavelet-based normalization method is proposed.Initially a face image is decomposed into its low frequency and high frequency components by 3-level wavelet transform,and then a histogram equalization is performed on the low frequency components sensitive to illumination,which can reduce the effect of illumination variation.At the same time,the detail（high frequency） components are adaptively de-noised and then accentuated by multiplying by a scalar to enhance edges.A normalized image is obtained from the modified components by inverse wavelet transformation.Experimental results on the Yale face database B show that the proposed method can effectively reduce the effect of illumination variation on face recognition,and can significantly improve the performance of face recognition system.