基于费歇尔脸方法的人脸识别算法  

Face Recognition Algorithms Based on Fisherface Method

作　　者：叶富东[1] 张富铭[2] 

机构地区：[1]湖北生态工程职业技术学院现代教育技术中心,武汉430200 [2]武汉理工大学理学院数学系,武汉430070

出　　处：《计算机与数字工程》2009年第1期138-139,191,共3页Computer & Digital Engineering

摘　　要：提出一种基于费歇尔方法的人脸识算法。分别论述了特征脸和费歇尔脸方法用于人脸识别的基本原理,给出了详细的算法过程。最后在ORL人脸数据库对算法进行了验证,采用三阶近邻分类器对测试图像进行了分类识别,比较了两种算法的识别效率。实验表明,费歇尔方法方法要优于特征脸方法。This paper proposes a face recognition algorithm based on Fisherface method. The theory of face has discussed recognition based on Eigenface and Fisherface has discussed respectively, and gave all the details of algorithm process. Finally, validated the algorithms on the ORL Database is validated. The test images were classified and identified by third-order neighbor classifier, and the recognition efficiency of the two algorithms were compared. Experimental results show that Fisherface method is much better than Eigenface method.

关 键 词：人脸识别 EIGENFACE FISHERFACE 特征提取三阶近邻 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]