《K近邻和最小二乘支持向量机相融合的人脸识别.》.pdf.pdf

刘红健:K近邻和最小二乘支持向量机相融合的人脸识别
《激光杂志》2014年第35卷第11期 LASER JOURNAL() 23
K近邻和最小二乘支持向量机相融合的人脸识别
刘红健 1,胡蓉 2
,广州 510725;
(,广州 510630 )
摘要:为了获得更加理想的人脸识别结果,提高人脸识别正确率,提出一种 K 近邻和最小二乘支持向量机
相融合的人脸识别方法(KNN-LSSVM)。首先采集人脸图像,提取人脸图像特征,并采用 KNN 删除特征向量中
的重复特征,得到人脸图像的特征向量;然后将特征向量输入到最小二乘支持向量机训练,建立相应的人脸分
类器;最后采用 ORL 人脸数据库和 Yale 人脸库进行仿真实验。仿真结果表明,KNN-LSSVM 提高了人脸识别
的正确率和识别效率,且具有较强的鲁棒性。
关键词:人脸识别;提取特征;最小二乘支持向量机;人脸分类器
中图分类号:TP391 文献标识码:A DOI编码:.
Face recognition based on K nearest neighbor and least square support vector machine
LIU Hong-jian1, HU Rong2
1. Department of art design, Guangzhou Maritime Institute, guangzhou 510725, China;
(2. Department puter, South China Normal University , guangzhou 510631, China)
Abstract:In order to obtain good face recognition results and improve the face recognition rate, this paper pro-
posed a new face recognition method based on K nearest neighbor and least square support vector machine. Firstly,
face images are collected and features are extracted, and the features are selected by K nearest neighbor algorithm,
and then the features are input to least square support vector machine to train and established image classifier, finally
the simulation experiment is carried out on the ORL face database and Yale face database. The simulation results
show that the proposed met