基于广义夹角的遥感图象计算机分类方法

Computer classification of generalized angle-based remote sensing image

通过研究遥感 ( RS)图象的物理基础 ,分析了传统分类方法的弊端 ,指出基于广义夹角的计算机分类方法 ( SAM法 )较适合地物光谱的形状特征 ,进而对 SAM法进行研究 ,并提出了改进方法和实施步骤 ,给出了分析比较试验实例 。

This paper analyzes tradition classification absence by studying physical base of RS image,points out the SAM method suits to the shape feature of surface spectrum,and studies the improving way and implement step,then it gives out an example about analysis and comparison experiment,which shows that improving method makes classification more accurate.