基于H/α/A分解的PolSAR农作物分类方法研究

Study on PolSAR Classification Based on H/α/A Decomposition for Crops

深入讨论了基于H-α分割的Wishart非监督分类方法在农作物分类中的性能;为改善农作物分类效果,引入了极化相关矩阵的最大特征值(Lam),并对AIRSAR数据进行了基于H-Lam分割平面的Wishart迭代分类处理。实验结果表明,与H/-αWishart非监督分类相比,该方法提高了农作物的分类效果。

The performance of Wishart classification for crops based on H-α segmentation plane is discussed in detail firstly. Then a mass of polarimetric feature parameters is analyzed with a purpose for improving crop classification performance, and the maximum eigenvalue （Lain） of polarimetric coherency matrix is introduced, and finally crop Wishart classification is carried out on AIRSAR data based on the H-Lam segmentation plane. Experimental result shows that the proposed method improves the effect of crop classification.