Significant employment of polarimetric synthetic aperture radar (SAR) is the classification of the Earth’s surface, which includes several processing algorithms, labelling, and representation options. This paper investigated the potential performance of the polarimetric reference function in SAR data processing to apply in coherent decomposition. The proposed methodology employed the simulated radar cross section of dihedral and trihedral corner reflector under a SAR geometry using physical optics (PO) and geometrical optics (GO) to produce the polarimetric reference function for azimuth compression in the processing of SAR data. The Pauli and the Krogager decompositions processed the compressed SAR data to investigate the effect and performance of the polarimetric reference function. The colour clustering results based on k-mean clustering algorithm could deliver more related scattering characteristic. The image pixel classification was strongly associated with the polarimetric reference function in processing step. These findings indicated that the obtained decomposition and clustering results could provide a better performance in visual impact and quantitative evaluation according to the decomposition term, as well as its physical interpretation.

中文翻译：

---

相干目标分解处理SAR数据极化参考函数的潜在性能

极化合成孔径雷达 (SAR) 的重要用途是对地球表面进行分类，其中包括多种处理算法、标记和表示选项。本文研究了极化参考函数在 SAR 数据处理中的潜在性能，以应用于相干分解。所提出的方法利用物理光学 (PO) 和几何光学 (GO) 在 SAR 几何结构下模拟二面角和三面角反射器的雷达截面，以在处理 SAR 数据时生成用于方位压缩的极化参考函数。Pauli 和 Krogager 分解处理压缩的 SAR 数据以研究极化参考函数的效果和性能。基于k-mean聚类算法的颜色聚类结果可以提供更多相关的散射特征。图像像素分类与处理步骤中的极化参考函数密切相关。这些发现表明，根据分解项及其物理解释，获得的分解和聚类结果可以在视觉冲击和定量评估方面提供更好的性能。