摘要
当混合像元分解将遥感中的分类问题深入到亚像元级别时就有巨大的研究前景。围绕混合像元分解问题,介绍了不同的端元提取方法,并利用地面真实遥感数据,采用PPI与SMACC 2种算法进行端元提取,完成丰度反演,获取每种端元的比例。对精度进行对比和分析的结果表明,在端元提取上PPI算法更为精确。
Mixed pixels immixing expand classification problem to a sub-pixel-level,it has a broad prospect in theory.It focus on the immixing of mixed pixels,describes several algorithms of end member extraction.Use the remote sensing image to do experiment,it getting the proportion of each end member.Finally,compare and analysis the different accuracy,then come to the conclusion that the PPI algorithm is more accurate.
出处
《现代测绘》
2016年第1期11-13,20,共4页
Modern Surveying and Mapping
关键词
混合像元
端元提取
线性光谱混合模型
mixed Pixel
end member extraction
linear spectral mixture model
