Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Science of Surveying and Mapping》 2015-01
Add to Favorite Get Latest Update

Comparison among four land cover change detection methods using pixel spectral information

ZHOU Xiao-guang;ZENG Lian-bing;YUAN Yu-cai;SONG Zheng-xiang;LI Fei;School of Geosciences and Info-Physics,Central South University;  
There are several change detection methods for remote sensing imagery.Among these methods,differencing,NDVI(Normalized Difference Vegetation Index),ratioing,and PCA(Principal Components Analysis),are the most popular change detection methods that utilize the pixel-by-pixel spectral information.Thus the comparison among these four methods was done from the updating perspective in this paper.The accuracy of each detection method was analyzed at first;then the sameness and difference among the result imagery of the four detection methds were presented;the accuracy of the reasonable combination methods was explored for the accuracy change detection of land cover using 30 mresolution imagery at last.The experiments showed that some reasonable combination methods could provide better change detection result than any single method for the change detection of land cover.The all changed pixels detected by these four methods were included in the result imagery of NDVI union PCA union ratioing,and the producer's accuracy achieved the best in the experiment.Therefore,the combination method union NDVI,PCA and ratioing would be the most appropriate method for the land cover change detection using 30 m TM imagery in updating application.
【Fund】: 国家高技术研究发展计划(863计划)重点项目(2009AA122004);; 国家自然科学基金项目(41371366);; 十二五科技支撑计划项目(2012BAK12B01-2)
【CateGory Index】: TP751
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved