ALOS PALSAR全极化图像分类研究

Research on Full Polarization Image Classification of ALOS PALSAR

极化干涉相干矩阵服从复Wishart分布,通过对相关系数的分析可以获得不同的地物类别。在总结极化干涉非监督Wishart ML分类流程的基础上,基于该方法对塔河地区全极化PALSAR数据进行了分类,研究结果表明:基于极化干涉的分类方法能够有效区分不同散射机制对应的地物,该分类方法具有较强的适应性,并且类间边界比较明显,这些分类信息为森林资源的开发和利用提供了参考。

The coherence matrix of polarization interference obeys the complex Wishart distribution, through the analysis of the correlation coefficient can get the categories of different features. Based on the summarizes of the process of polarization interference unsupervised Wishart ML classification, the full polarization PALSAR data of Tahe region were classified by this method, the study results show that the classification method based on polarization interference can effectively distinguish corresponding features of different scattering mechanisms, the classification method has strong adaptability, and the boundaries between the categories are more obvious, and the category information provide a reference for the development and utilization of forest resources.