论文部分内容阅读
本文结合Freeman分解和子孔径分析,提出一种新的极化SAR图像分类算法。该方法首先利用子孔径分解,产生不同方位观察角度下的子孔径图像,再利用Freeman分解对各个子孔径图像提取三种散射机理成分的功率,平均后对类别进行细分,最后使用Wishart统计分类器对类别进行分类划分得到最终结果。该方法考虑了极化散射机理在不同方位观察角度下的变化,能够取得较好的分类效果,能够保存主要极化散射特性的纯度,同时还可以动态地设定分类类别数。最后利用EMISAR获取的极化SAR数据进行了仿真,验证了该方法的有效性。
In this paper, a new Polarimetric SAR image classification algorithm is proposed based on Freeman decomposition and sub-aperture analysis. Firstly, the sub-aperture image is generated by using sub-aperture decomposition, and the sub-aperture images are generated at different azimuth observation angles. Then, the power of three scattering mechanism components is extracted from each sub-aperture image by Freeman decomposition, and the classification is subdivided according to the average by Wishart statistical classification The classification of categories by category to get the final result. The method takes into account the change of polarimetric scattering mechanism under different azimuth observation angles, achieves good classification results, preserves the purity of the main polarization scattering characteristics, and simultaneously sets the number of classification categories. Finally, the polarization SAR data acquired by EMISAR are simulated and the effectiveness of this method is verified.