摘要: 主分量分析(Principal Component Analysis,PCA)是模式识别领域中一种重要的特征抽取方法,该方法通过K-L展开式来抽取样本的主要特征。基于此,提出一种拓展的PCA人脸识别方法,即分块排序PCA人脸识别方法(MSPCA)。分块排序PCA方法先对图像矩阵进行分块,对所有分块得到的子图像矩阵利用PCA方法求出矩阵的所有特征值所对应的特征向量并加以标识;然后找出这些所有的特征值中k个最大的特征值所对应的特征向量,用这些特征向量分别去抽取所属的子图像的特征;最后,在MSPCA的基础上,将抽取子图像所得到的特征矩阵合并,把这个合并后的特征矩阵作为新的样本进行PCA+LDA。与PCA和PCA+LDA方法相比,分块排序PCA由于使用子图像矩阵,可以避免使用奇异值分解理论,从而更加简便。在ORL人脸库上的实验结果表明,所提出的方法在识别性能上明显优于经典的PCA和PCA+LDA方法。
[1] Belhumeur P N,Hespanha J P,Kriengman D J.Eigenfaces vsFisherfaces:Recognition using class specific linearprojection [J].IEEE Trans on Pattern Anal Machine Intell,1997,9(7):711-720 [2] 边肇祺,张学工.模式识别(第二版)[M].北京:清华大学出版社,1999:176-177 Bian Zhao-qi,Zhang Xue-gong.Pattern Recognition [M].Beijing:Tsinghua University Press,2000 [3] Kirby M,Sirovich L.Application of the KL procedure for the characterization of human faces [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1990,2(1):103-108 [4] Turk M,Pentland A.Face processing:Models for recognition[C]∥Proceedings of Intelligent Robots and Computer Vision VIII.1989,1:22-32 [5] Turk M,Pentland A.Eigenfaces for recognition [J].Journal of Cognitive Neuroscience,1991,3(1):71-86 [6] Turk M,Pentland A.Face recognition using Eigenfaces [C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Hawaii,USA,1991:586-591 [7] 金忠.人脸图像特征抽取与维数研究 [D].南京:南京理工大学,1999 Jin Zhong.Feature extraction and dimension research of human faces[D].Nanjing:Nanjing University of science and technology,1999 [8] Yang Jian,Yu Yang-jing.Why can LDA be performed in PCA transformed space[J].Pattern Recognition,2003,6:563-566 [9] 杨键,杨静宇,叶辉,等.Fisher线性鉴别分析的理论研究及其应用[J].自动化学报,2003,9(4):482-493 Yang jian,Yang jing-yu,Ye Hui,et al.Theory of fisher linear discriminant analysis and its apllication[J].Acta Automatica Sinica,2003,9(4):482-493 [10] Jin Zhong,Yang Jing,et al.Face Recognition based on uncorrelated discriminant transformation[J].Pattern Recognition,2001,4(7):1405-1416 [11] Chen Li-fen,Liao Hong-yuan,et al.A new LDA-Based FaceRecognition System Which Can Solve the Small Sample Size Problem [J].Pattern Recognition,2000,3:1713-1726 [12] Wu Xiao-jun,Josef K,Yu Yang-jing.A New Direct LDA Algorithm for Feature Extraction in Face Recognition [C]∥Procee-dings of the 17th International Conference on Pattern Recognition.Cambridge,UK,2004:545-548 [13] 陈伏兵,高秀梅,张生亮,等.基于分块 PCA的人脸识别方法[J].小型微型计算机系统,2006,7(10):1943-1947 Chen Fu-bing,Gao Xiu-mei,Zhang Sheng-liang,et al.Face Recognition based on modular PCA approach[J].Mini-Micro Systems,2006,7(10):1943-1947 [14] 陈伏兵,杨静宇.分块PCA及其在人脸识别中的应用[J].计算机工程与设计,2007,8(8):1889-1892 Chen Fu-bing, Yang Jing-yu.Modular PCA and its application in face recognition[J].Computer Engineering and Design,2007,8(8):1889-1892 |
No related articles found! |
|