首页 | 本学科首页   官方微博 | 高级检索  
     

基于线性混合模型的高光谱图像端元提取
引用本文:薛绮,匡纲要,李智勇.基于线性混合模型的高光谱图像端元提取[J].遥感技术与应用,2004,19(3):197-201.
作者姓名:薛绮  匡纲要  李智勇
作者单位:国防科学技术大学电子科学与工程学院一系,湖南,长沙,410073
摘    要:近年来,基于线性混合模型的光谱解混合技术正在越来越广泛地用在光谱数据分析和遥感地物量化中,这项技术的关键就在于确定端元(Endmember)光谱。通常,端元的荻取有两种方式:来源于光谱库以及来源于图像数据,相比之下后者得到的结果更能体现真实的地面信息。为此,从线性混合模型的特点出发,归纳了目前几种比较成熟的端元提取算法,分析了它们的主要思想和存在的优缺点,并总结了评估算法结果的依据,最后介绍了端元提取技术的发展趋势。

关 键 词:高光谱  端元提取  线性混合模型
文章编号:1004-0323(2004)03-0197-05
修稿时间:2003年11月27

Endmember Extraction Algorithms from Hyperspectral Image Based on The Linear Mixing Model:An Overview
XUE Qi,KUANG Gang-yao,LI Zhi-yong.Endmember Extraction Algorithms from Hyperspectral Image Based on The Linear Mixing Model:An Overview[J].Remote Sensing Technology and Application,2004,19(3):197-201.
Authors:XUE Qi  KUANG Gang-yao  LI Zhi-yong
Abstract:Recently, spectral unmixing techniques based on the linear mixing model are widely used for hyperspectral data analysis. The identification of the purest spectral signatures of ground constituents (endmember) in a scene is a previous step in all unmixing approaches. Once these spectra are found, the image cube can be "unmixed" into fractional abundances of each material in each pixel. Compared with the pure spectral from libraries, the spectra from the image are adapt to spectral mixture analysis. Linear spectral unmixing is one of the most important approaches for the analysis and classification of hyperspectral datasets. So several different algorithms for extracting endmembers are derived from the linear mixing model, including the most successful Pixel Purity Index (PPI); ORASIS, the first real time autonomous algorithm; N-FINDR, an algorithm that determines endmembers by inflating a simplex; Iterative Error Analysis (IEA) which finds endmembers by iterative unmixing,and Automated Morphological Endmember Extraction (AMEE) that considers the existing spatial correlation between pixels. Then the main ideas of these methods, as well as the advantages, disadvantages and applications are presented and discussed respectively in this paper. After that, the criterions for evaluating the quality of selected endmembers are summarized, the trends of endmember extraction algorithms in recent years are also analyzed.
Keywords:Hyperspectral  Endmember extraction  The linear mixing model
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号