基于H-α和改进C-均值的全极化SAR图像非监督分类

该文提出一种基于H-α和改进C-均值的全极化SAR图像非监督分类方法.该方法先按H-α对全极化SAR图像进行基于散射机理的分类,再将分类结果作为改进C-均值算法的初始类别划分,从而实现地物分类.迭代次数确定是C-均值动态聚类算法的关键,文中利用图像熵给出了一种新的迭代终止准则.与H-α方法相比,该文方法能在保留分类结果物理散射机理的同时,实现有效的地物分类.NASA/JPL实验室AIRSAR系统获取的L波段旧金山全极化SAR数据的实验结果验证了该文方法的有效性.

Unsupervised Classification of Fully Polarimetric SAR Image Using H-α Decomposition and Modified C-Mean Algorithm

A new method for unsupervised classification of terrain types using fully POLarimetric Synthetic Aperture Radar (POLSAR) data is proposed in this paper. The method is a combination of the unsupervised classification based on Cloude’s H-α decomposition and the modified C-mean algorithm. The fully polarimetric SAR image is initially classified using Cloude’s method. The classification map is used as input of the modified C-mean algorithm, and then iteration is performed. It is important to determine the number of iteration in the modified C-mean algorithm, and a new termination criterion is presented using image entropy to do so. Compared with H-α decomposition, not only scattering mechanisms of all classes can be preserved, but also terrain classification is effectively performed using this method. The effectiveness of this method is demonstrated using an L-band fully polarimetric SAR image of San Francisco, acquired by the NASA/JPL AIRSAR sensor.