基于小波的人脸光照补偿技术研究

Improving Variable Illumination on Wavelet

在人脸识别系统中，人脸光照补偿是预处理模块最关键的部分。采集到的人脸图像往往因为光照方向、强度等的不同而使得一张图片中的人脸光照不均匀，降低了定位精度和识别率。该文提出了基于小波的人脸光照补偿，通过MAT-LAB7．1平台对Yale人脸库中的图像进行实验，能达到很好的效果。

In the person face recognition system, the person face illumination compensation is the most essential part in the pre- processing module. Because of the different direction and intensity of illumination, the gathering person face image＇s illumination is not uniform, which can reduce the pointing accuracy and the recognition rate. This article uses a way based on the wavelet for illumination compensation, which can achieve a good effect by carrying on the experiment in Yale face storehouse through the MATLAB 7.1 platform.