%0 Journal Article
%T 关于二维主成分分析方法的研究
%A 王立威
%A 王潇
%A 常明
%A 封举富
%J 自动化学报
%P 782-787
%D 2005
%X ？Theprincipalcomponentanalysis(PCA),ortheeigenfacesmethod,isadefactostandardinhumanfacerecognition.NumerousalgorithmstriedtogeneralizePCAindifferentaspects.Morerecently,atechniquecalledtwo-dimensionalPCA(2DPCA)wasproposedtocutthecomputationalcostofthestandardPCA.UnlikePCAthattreatsimagesasvectors,2DPCAviewsanimageasamatrix.Withaproperlydefinedcriterion,2DPCAresultsinaneigenvalueproblemwhichhasamuchlowerdimensionalitythanthatofPCA.Inthispaper,weshowthat2DPCAisequivalenttoaspecialcaseofanexistingfeatureextractionmethod,i.e.,theblock-basedPCA.UsingtheFERETdatabase,extensiveexperimentalresultsdemonstratethatblock-basedPCAoutperformsPCAondatasetsthatconsistofrelativelysimpleimagesforrecognition,whilePCAismorerobustthan2DPCAinhardersituations.
%K Facerecognition
%K PCA
%K two-dimensionalPCA
%K block-basedPCA
%U http://www.aas.net.cn/CN/abstract/abstract15955.shtml