摘要
将光谱聚类方法应用于高光谱遥感数据处理,对低反射率地物信息的提取取得较好效果;同时采用决策树的多分类器组合方法提取高光谱遥感影像信息,经对比研究发现其效果明显优于单个分类器。
The spectral clumping method is applied to deal with hyperspectral remote sensing data. It is very well to extract information of low reflectance objects. At the meanwhile, it is also applied multiple classifiers combination methods based on decision tree to extracting information from hyperspeetral remote sensing images. It is discovered that multiple classifiers method is clearly superior to single classifier in the research.
出处
《高原地震》
2009年第4期40-45,共6页
Plateau Earthquake Research
关键词
高光谱遥感技术
多分类器
决策树
Hyperspeetral remote sensing technology
Multiple classifiers
Decision tree
