摘要
为了提高人脸正确识别率和效率,在行列方向的二维线性判别分析((2D)2LDA)基础之上,提出了一种二维复判别分析(2DCCDA)的人脸识别方法。该方法通过(2D)2LDA并行提取到的行和列特征矩阵,利用复二维鉴别式分析(C2DLDA)将行和列特征融合成复数特征矩阵,从复数特征矩阵中提取出最具分类能力的系数组成特征向量。相比较二维线性判别分析(2DLDA)和(2D)2LDA方法,2DCCDA需要更少的特征系数来表征一幅图像,并且正确识别率也相应提高。
To increase the accurate recognition rate and the efficiency of face recognition,on the basis of two-dimensional linear discriminant analysis（2DLDA） along row and column directions（（2D）2LDA）,a face recognition method is proposed,based on two-dimensional combined complex discriminant analysis（2DCCDA）.In this method,firstly row and column feature matrixes are extracted by（2D）2LDA.Secondly,both feature matrixes are integrated into a complex feature matrix using complex version of 2DLDA（C2DLDA）.In the end,some components are picked out from the complex feature matrix to form a feature vector according to their discriminative abilities.Compared with 2DLDA and（2D）2LDA methods,the proposed 2DCCDA can present an image with less feature components and the accurate recognition rate is improved.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第11期2514-2518,共5页
Computer Engineering and Design
基金
广东省教育部科技部企业科技特派员行动计划专项基金项目(2009B090600034)
广东省科技计划基金项目(2009B060700124)
广州市教育科学"十一五"规划基金项目(07B171)
关键词
人脸识别
主成份分析
线性判别分析
复二维鉴别式分析
二维复判别分析
face recognition
principal component analysis（PCA）
two-dimensional linear discriminant analysis
complex version of 2DLDA
two-dimensional combined complex discriminant analysis
