• • Previous Articles     Next Articles

PIXEL DECOMPOSITION BASED ON PRINCIPAL COMPONENT ANALYSIS

Shao Meizhen Deng Guanglin Information Engneering Institute of the PLA,Zhengzhon) Xu Xiru (Institute of Remote Seneing,Peking University)   

  • Published:1989-10-25

Abstract: In this paper,the method of mixed pixel decomposition based onPrincipal Component Analysis(PCA)is mainly introduced.Usually PCA isapplied to enhance image and reduce the dimensions of data space,withoutchanging the space construction of data.With the method of target test in thispaper,it can be realized to decompose mixed pixels.Because of the fixedresolution of Land-sat data,a large amount of mixed pixels exists in remotesensing image.Is one of the major reasons of the same spectrum for differentsurface features and different spectrums for the same surface feature.If it ispossible to decompose mixed pixels,this will be very helpful for extensivelyutilizing the remote sensing data.