改进Retinex和稀疏表示的光照人脸识别

为了提高光照变化条件下的人脸识别率，针对当前人脸识别方法存在的缺陷，提出了一种改进Retinex算法和稀疏表示相融合的光照人脸识别方法。首先对Retinex算法的不足进行改进，并应用于人脸图像预处理中，消除光照对人脸识别的干扰，然后采用稀疏表示提取人脸特征向量，并采用投票方式实现人脸识别，最后通过3个标准人脸数据库对方法的性能进行测试。结果表明，该方法不仅提高了人脸识别率，而且缩短了人脸识别时间，对光照具有较好的鲁棒性。

Face recognition under illumination variation condition based on improved Retinex and sparse representation

In order to improve the face recognition under illumination variation condition, a novel face recognition algorithm based on improved Retinex and sparse representation is proposed in this paper to the defects of face recognition algorithms. Firstly, Retinex algorithm is improved and applied to the face image preprocessing to eliminate the illumination variation interference in face recognition, and then the sparse representation is used to extract face feature vector and vote method is used to recognize face, and finally the performance of algorithm is tested on three face data set. The results show that the proposed algorithm can not only improve effectively the face recognition rate, but also shorten the time of face recognition, and it has good robustness for illumination variation.