| 改进的广义主分量分析及在人脸识别中的应用 |
| |
| 引用本文： | 王建国,林宇生,刘锁兰,杨静宇.改进的广义主分量分析及在人脸识别中的应用[J].计算机工程与设计,2008,29(11):2901-2902. |
| |
| 作者姓名： | 王建国 林宇生 刘锁兰 杨静宇 |
| |
| 作者单位： | 1. 南京理工大学,计算机科学与技术学院,江苏,南京,210094;唐山学院,网络教育中心,河北,唐山,063000
2. 南京理工大学,计算机科学与技术学院,江苏,南京,210094 |
| |
| 摘 要： | 广义主分量分析是一种利用图像矩阵直接计算的二维主分量分析,较传统主分量分析提高了特征抽取速度及识别率.通过对广义主分量分析中的产生矩阵进行分析,并重新定义,在类间散布矩阵定义的基础上引入了径向基函数,通过调整径向基函数的系数得到更有利于分类的特征信息,获得较高的识别率.在Yale,ORL两个人脸数据库上的实验结果表明了改进方法的有效性和鲁棒性.
|
| 关 键 词： | 主分量分析 广义主分量分析 图像矩阵 散布矩阵 特征抽取 人脸识别 |
| 文章编号： | 1000-7024(2008)11-2901-02 |
| 修稿时间： | 2007年6月8日 |
Improved generalized principal component analysis and its application to face recognition |
| |
| WANG Jian-guo,LIN Yu-sheng,LIU Suo-lan,YANG Jing-yu.Improved generalized principal component analysis and its application to face recognition[J].Computer Engineering and Design,2008,29(11):2901-2902. |
| |
| Authors: | WANG Jian-guo LIN Yu-sheng LIU Suo-lan YANG Jing-yu |
| |
| Affiliation: | WANG Jian-guo1,2,LIN Yu-sheng1,LIU Suo-lan1,YANG Jing-yu1 (1. School of Computer Science , Technology,Nanjing University of Science , Technology,Nanjing 210094,China,2. Network , Education Center,Tangshan College,Tangshan 063000,China) |
| |
| Abstract: | Generalized PCA (GPCA) is a method directly based on original image matrices, and its eigenvectors are derived for image feature extraction, and the speed of feature extraction is faster than CPCA. The generation matrix of the GPCA is analyzed, and the between-class scatter matrix is redefined by introducing a radical basis function, so classification features are obtained by adjusting the coefficient of the function. Experiments are conducted on Yale and ORL face databases, and the experimental results ind... |
| |
| Keywords: | principal component analysis generalized principal component analysis image matrix scatter matrix feature extraction face recognition |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
