SAM权重法在端元提取中的应用  

Application of SAM in Endmember Extraction

作　　者：朱怀朝[1] 王茂芝[1] 崔之熠[1] 刘国涛[1] 芶盛[1] 

机构地区：[1]成都理工大学管理科学学院数学地质四川省高校重点实验室,成都610059

出　　处：《四川理工学院学报（自然科学版）》2011年第2期214-217,共4页Journal of Sichuan University of Science & Engineering(Natural Science Edition)

基　　金：国家863计划重点课题(2008AA121103)

摘　　要：高光谱遥感的出现使得在宽波段遥感中不可探测的物质能被探测。获得的高光谱数据大都具有数据冗余度高、信噪比低等特点。文章通过idl编程实现高光谱数据的特征提取并利用其做了端元提取流程与光谱解混,及权重法SAM端元提取、混合光谱分解模型及实现。利用SAM权重法能够完成端元提取并最终得到的解混结果。The emergence of hyperspectral remote sensing technology makes the undetected interested object can be detected by remote sensing technology today,it is important for extracting data with features as high redundancy and low signal-to-noise ratio of hyperspectral data.It contains space,radiation and spectrum.The paper analyzes the theory with IDL to obtain feature extraction,and uses the results to obtain endmember extraction and spectral unmixing algorithm and dpectral angle mapper.Through this test,the technology and method of endmember extraction and spectral unmixing in hyperspectral remote sensing is taking shape.

关 键 词：高光谱遥感 像元纯净指数 特征提取 光谱角度填图 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]