一种基于加权变形的2DPCA的人脸特征提取方法

曾岳; 冯大政

doi:10.3724/SP.J.1146.2010.01003

一种基于加权变形的2DPCA的人脸特征提取方法

doi: 10.3724/SP.J.1146.2010.01003

曾岳,
冯大政

基金项目:

国家自然科学基金(60372049)和江西省科技计划青年基金(GJJ09412)资助课题

计量
- 文章访问数: 3703
- HTML全文浏览量: 137
- PDF下载量: 1096
- 被引次数: 0
出版历程
- 收稿日期: 2010-09-14
- 修回日期: 2010-11-29
- 刊出日期: 2011-04-19

An Algorithm of Feature Extraction of Face Based on the Weighted Variation of 2DPCA

摘要

摘要: 该文首先分析了主成分分析法(PCA)和2维主成分分析法(2DPCA)的关系，针对2DPCA丢失具有鉴别能力的协方差信息以及PCA方法不能解决小样本的问题，提出了基于一种加权变形的2DPCA的人脸特征提取方法(WV2DPCA)，该方法利用变形的2DPCA方法分别对人脸3个子部分分别提取特征，然后根据最近邻理论和权值进行分类。经过在ORL人脸库和YALE人脸库的实验研究表明：与2DPCA相比，提高了人脸空间的识别率，压缩了人脸空间的系数，减少了识别时间；在识别的准确率方面，更优于传统的Fisherfaces，IC，Kernel Eigenfaces的算法。
- 人脸识别 /
- 人脸表示 /
- 主成分分析法(PCA) /
- 2维主成分分析法(2DPCA)
Abstract: This paper first discusses the relationship of Principal Component Analysis (PCA) and 2-Dimensional PCA (2DPCA). For 2DPCA eliminating the some covariance information which can be useful for recognition, and PCAs small sample size problem, an algorithm of feature extraction of face based on the Weighted Variation 2DPCA (WV2DPCA) is proposed. Three sub-parts of the face features are extracted respectively in the method of the variation of 2DPCA, and then are classified according to the weight and the Nearest neighbor theory. The experiments on both of ORL and YALE face bases show improvement in recognition accuracy, fewer coefficients and recognition time over 2DPCA, and this algorithm is also superior to the traditional eigenfaces, ICA and Kernel Eigenfaces in terms of the recognition accuracy.
- Face recognition /
- Face representation /
- Principal Component Analysis (PCA) /
- 2-Dimensional PCA (2DPCA)

HTML全文

参考文献(1)

Yang J, Zhang D, and Alejandro F, et al.. Two-dimensional PCA: a new approach to appearance-based face representation and recognition. IEEE Transaction on Pattern Analysis and Machine Intelligence, 2004, 26(1): 131-137.[2] Rajkiran G and Asari V K. An improved face recognition technique based on modular PCA approach. Pattern Recognition Letters, 2004, 25(4): 429-436.[3] Zuo Wang-meng, Zhang D, and Wang Kuan-quan.Bidirectional PCA with assembled matrix distance Metric for image recognition. IEEE Transaction on Systems, Man, and Cybernetics-part B: Cyebrnetics, 2006, 36(4): 863-872.[4] Chen Song-can and Zhu Yu-lian. Subpattern-based principle component analysis. Pattern Recognition, 2004, 37(5): 1081-1083.[5] Tan Ke-ren and Chen Song-can. Adaptively weighted sub-pattern PCA for face recognition. Neurocomputing, 2005, 64(1): 505-511.[6] Eftekhari A. Mohamad Forouzanfar and Hamid Abrishami Moghaddam. Block-wised 2D kernel PCA/LDA for recognition. Information Processing Letters, 2010, 110(17): 761-766.[7] Huang Guo-hong. Fusion (2D)2PCALDA: a new method for face recognition. Applied Mathematics and Computation, 2010, 216(11): 3195-3199.[8] Qi Yong-feng and Zhang Jia-shu. (2D)2PCALDA: an efficient approach for face recognition. Applied Mathematics and Computation, 2009, 213(1): 1-7.[9] Wang Jin, Barret A, and Wang Lu, et al.. Multilinear principal component analysis for face recognition with fewer features. Neurocomputing, 2010, 73(10-12): 1550-1555.[10] Yu W, Wang Z, and Chen W. A new framework to combine vertical and horizontal information for face recognition. Neurocomputing, 2009, 72(4-6): 1084-1091.[11] Zheng Wei-shi, Lai J H, and Li S Z. 1D-LDA vs. 2DLDA: when is vector-based linear discriminant analysis better than matrix-based?. Pattern Recognition, 2008, 41(7): 2156-2172.[12] Belhumeur P N, Hespanha J P, and Kriengman D J. Eigenfacs vs. fisherfaces: recognition using class specific linear projection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997, 19(7): 711-720.[13] Yuen P C and Lai J H. Face representation using independent component analysis. Pattern Recognition, 2002, 35(6): 1247-1257.[14] Yang M H. Kernel eigenfaces vs. kernel fisherfaces: face recognition using kernel methods. Proc Fifth IEEE int1 conf. Automatic Face and Gesture Recognition(RGR02), Washinton D C, May 2002: 215-220.

施引文献

资源附件(0)

访问统计