首页 | 本学科首页   官方微博 | 高级检索  
     

基于核的Fisher非线性最佳鉴别分析在人脸识别中的应用
引用本文:成新民,蒋云良,胡文军,吴小红.基于核的Fisher非线性最佳鉴别分析在人脸识别中的应用[J].中国图象图形学报,2007,12(8):1395-1400.
作者姓名:成新民  蒋云良  胡文军  吴小红
作者单位:湖州师范学院信息工程学院,湖州师范学院信息工程学院,湖州师范学院信息工程学院,湖州师范学院信息工程学院 湖州313000,湖州313000,浙江大学计算机科学与技术学院,杭州310027,湖州313000,湖州313000
基金项目:国家自然科学基金;浙江省自然科学基金
摘    要:抽取最佳鉴别特征是人脸识别中的重要一步。对小样本的高维人脸图像样本,由于各种抽取非线性鉴别特征的方法均存在各自的问题,为此提出了一种求解核的Fisher非线性最佳鉴别特征的新方法,该方法首先在特征空间用类间散度阵和类内散度阵作为Fisher准则,来得到最佳非线性鉴别特征,然后针对此方法存在的病态问题,进一步在类内散度阵的零空间中求解最佳非线性鉴别矢量。基于ORL人脸数据库的实验表明,该新方法抽取的非线性最佳鉴别特征明显优于Fisher线性鉴别分析(FLDA)的线性特征和广义鉴别分析(GDA)的非线性特征。

关 键 词:人脸识别  Fisher非线性鉴别分析  核方法  小样本问题  病态问题
文章编号:1006-8961(2007)08-1395-06
修稿时间:2004-04-292006-07-18

Face Recognition Based on Kernel Fisher Nonlinear Optimal Discriminant Analysis
CHENG Xin-min,JIANG Yun-liang,HU Wen-jun,WU Xiao-hong,CHENG Xin-min,JIANG Yun-liang,HU Wen-jun,WU Xiao-hong,CHENG Xin-min,JIANG Yun-liang,HU Wen-jun,WU Xiao-hong and CHENG Xin-min,JIANG Yun-liang,HU Wen-jun,WU Xiao-hong.Face Recognition Based on Kernel Fisher Nonlinear Optimal Discriminant Analysis[J].Journal of Image and Graphics,2007,12(8):1395-1400.
Authors:CHENG Xin-min  JIANG Yun-liang  HU Wen-jun  WU Xiao-hong  CHENG Xin-min  JIANG Yun-liang  HU Wen-jun  WU Xiao-hong  CHENG Xin-min  JIANG Yun-liang  HU Wen-jun  WU Xiao-hong and CHENG Xin-min  JIANG Yun-liang  HU Wen-jun  WU Xiao-hong
Affiliation:1. School of Information Engineering, Huzhou Teachers College, Huzhou 313000;2. College of Computer Science, Zhefiang University, Hangzhou 310027
Abstract:Extracting the most discriminatory features is important in face recognition tasks. In the case of a small number of face samples, as the existed methods for extracting nonlinear most discriminatory face features encounter various problems. So a new method for extracting fisher nonlinear most discriminatory features is proposed in this paper. The fisher criterion is formulated using between-class scatter matrix and within-class scatter matrix based on kernel method. Thus nonlinear most discriminatory features are obtained. However, this method causes ill-problem. To solve this problem, we search optimal discriminant vectors in null space of within-class scatter matrix. Repeated experimental results on ORL database indicate that the proposed method significantly outperforms the Fisher linear discriminant analysis(FLDA) and generalized discriminant analysis(GDA).
Keywords:face recognition  Fisher nonlinear discriminant analysis  kernel method  small sample size problem  ill-pose problem
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《中国图象图形学报》浏览原始摘要信息
点击此处可从《中国图象图形学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号