摘要
ThispaperproposesaPCAandKPCAself-fusionbasedMSTARSARautomatictargetrecognitionalgorithm.Thisalgorithmcombinesthelinearfeatureextractedfromprincipalcomponentanalysis(PCA)andnonlinearfeatureextractedfromkernelprincipalcomponentanalysis(KPCA)respectively,andthenutilizestheadaptivefeaturefusionalgorithmwhichisbasedontheweightedmaximummargincriterion(WMMC)tofusethefeaturesinordertoachievebetterperformance.Thelinearregressionclassifierisusedintheexperiments.Theexperimentalresultsindicatethattheproposedself-fusionalgorithmachieveshigherrecognitionratecomparedwiththetraditionalPCAandKPCAfeaturefusionalgorithms.
出版日期
2012年04月14日(中国期刊网平台首次上网日期,不代表论文的发表时间)