基于KummerU和MRF的极化SAR分类算法研究

Study on Polarmetric SAR Classification Algorithm Based on Kummer U and MRF

为了加入空间关系和抑制斑点噪声,本文提出了基于Kummer U模型和马尔科夫随机场(MRF)的极化SAR分类算法,该算法采用Kummer U分布来建模极化SAR数据,根据Kummer U的参数变化能够建模各种类型的地物。同时,MRF框架加入图像的上下文关系,能够获得区域一致的分类结果。

A polarimetric SAR classification algorithm based on Kummer U model and MRF（ Markov Random Field） is presented to add space relation and reject spot noise. The algorithm utilizes Kummer U distribution to model polarimetric SAR data,and model various kinds of ground object on basis of parameters change of Kummer U. Meanwhile,MRF frame is added with context relation of image so as to obtain classification result with consistent area.