人脸识别中光照补偿参数及必要性的判定方法  

Compensated Illumination Direction Estimation Method in Face Recognition Based on 2DPCA and Fisher

作　　者：孙雪梅[1] 苏菲[1] 蔡安妮[1] 

机构地区：[1]北京邮电大学图像处理与生物特征识别实验室,北京100876

出　　处：《计算机应用研究》2007年第7期306-308,311,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(60472069)

摘　　要：针对实际人脸识别应用中大量人脸无须作光照补偿的情况,分别用2DPCA和Fisher两种人脸识别方法对FRVT标准人脸库进行识别,得出准确的影响人脸识别的光照方向参数,并提出了一种简单快速的判断是否需补偿的方法。实验证明,经过光照补偿判定后进行的人脸识别实验,需补偿人脸数目大量减少,运算量大大降低,人脸识别速度可以提高近四倍。To solve the problem that most of faces to recognize need not to be compensated in fact, this paper recognized faces in standard FRVT face database. Based on 2DPCA and Fisher methods respectively, it achieved the parameters on illumination influence to face recognition through the experiments. Meanwhile it proposed a simple and fast method to identify if the face under illumination should be compensated. The experimental results show that compensated faces and computation complexity are decreased greatly after compensation estimation. The recognition speed is up to four times than the traditional methods.

关 键 词：2DPCA方法 Fisher方法 光照影响参数 光照补偿 人脸识别 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]