全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

融合小波变换与KPCA的分块人脸特征抽取与识别算法

DOI: 10.11834/jig.20070413

Keywords: 小波变换,核主成分分析,分块人脸,特征抽取,支持向量机

Full-Text   Cite this paper   Add to My Lib

Abstract:

鉴于小波多尺度变换对高维图像特征具有良好的压缩和表达能力,提出了一种融合小波变换与KPCA(核主成分分析)方法的分块人脸特征抽取与识别算法。该算法首先对人脸图像进行分块小波变换,再根据图像块的位置分布选取不同的频率分量;然后对此分量进行KPCA特征抽取,并通过对抽取到的特征进行融合来得到最终人脸鉴别特征;最后利用支持向量机分类器进行特征分类与识别。通过对ORL和Yale标准人脸图像库的实验仿真结果表明,该算法不仅在识别性能和分类速度上明显高于传统的PCA方法及融合小波特征的KPCA方法,而且对于人脸光照、姿态和表情变化均具有良好的鲁棒性。

References

[1]  Turk M,Pentland A.Eigenfaces for recognition[J].Journal of Cognitive Neuroscience,1991,3(1):71 ~ 86.
[2]  Scholkopf B,Smola A,Muller K.Nonlinear component analysis as a kernel eigenvalue problem[J].Neural Computation,1998,10 (5):1299 ~ 1319.
[3]  Gao Xi-qi,Zhou Hong-xiang,He Zhen-yang.Face recognition based on wavelet transform and kernel principal component analysis[J].Journal of Southeast University,1996,26(2):137 ~ 141.[高西奇,周洪祥,何振亚.基于小波变换的主元分析人脸图象识别[J]东南大学学报,1996,26(2):137~141.]
[4]  Xie Yong-hua,Fu De-sheng,Ren Wei-hua.A fractal image coding algorithm research based on wavelet transformation[J].Journal of Image and Graphics,2003,7(8):839~842.[谢永华,傅德胜,任卫华.一种基于小波变换的图像分形编码压缩算法的研究[J].中国图象图形学报,2003,7(8):839~842.]
[5]  Chen Cai-kou,Wang Zheng-qun,Yang Jing-yu.A fast kernel principal component analysis technique based on wavelet feature[J].Computer Engineering and Applications,2004,22 (40):45 ~ 47.[陈才扣,王正群,杨静字.基于小波特征的快速核主分量分析技术[J].计算机工程与应用,2004,22(40):45~47.]
[6]  Weston J,Watkins C.Multi-class support vector machines[R].Technical Report CSD-TR-98-04 in Royal Holloway University of London,1998.
[7]  Belhumeur P N,Hespanha J P,Kriengman D J.Eigenfaces vs.fisherfaces:Recognition using class specific linear projection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1997,19(7):711 ~720.
[8]  Baudat G,Anouar F.Generalized discriminant analysis using a kernel approacn[J].Neural Computation,2000,12 (10):2385 ~2404.
[9]  Moghaddam B.Principal manifolds and probabilistic subspaces for visual recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24 (6):780 ~ 788.
[10]  Nastar C,Ayache N.Frequency-based non-rigid motion analysis[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1996,18 (11):1067 ~ 1079.
[11]  Zhou Guo-min,Chen Yong,Li Guo-jun.Two keys to the application of wavelet transform in human face recognition[J].Journal of Zhejiang University(Science Edition),2005,32(1):34 ~ 38.[周国民,陈勇,李国军.人脸识别中应用小波变换的两个关键问题[J].浙江大学学报,2005,32(1):34~38.]

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133