植被高度的极化干涉互协方差矩阵分解反演法

A Novel Vegetation Height Inversion Method Based on Polarimetric Interferometric Covariance Matrix Decomposition

经典三阶段极化干涉SAR植被高度反演算法中地面散射相位估计不准确,从而导致植被高度反演精度存在偏差。针对这一关键问题,提出基于极化干涉互协方差矩阵分解的植被高度反演新方法。该方法首先利用Freeman-Durden分解方法对极化干涉互协方差矩阵进行分解,从而估计出更准确的地面散射相位。然后,结合随机体-地表散射(RVoG)模型反演植被高度。最后,利用欧空局(ESA)的软件PolSARpro模拟的L波段极化干涉SAR数据和亚马逊森林地区的ALOS PALSAR L波段数据进行试验,结果表明本文算法提取的植被高度相比经典三阶段法精度更高,验证了算法的有效性和可靠性。

The results of vegetation height inversion based on classical three-stage method using polarimetric SAR interferometry data are seriously affected by the inaccurate estimation of underlying topographic phase.Therefore,this paper proposes a new inversion algorithm based on polarimetric interferometric covariance matrix decomposition.Firstly,the new method applies the Freeman-Durden decomposition concept to polarimetric interferometry covariance matrix decomposition for obtaining more accurate underlying topographic phase.Then,it combines random volume over ground(RVoG)coherent scattering model and the estimated underlying topographic phase to estimate vegetation height.Finally,the performance of the new inversion algorithm is demonstrated by using the simulated L-band PolInSAR data from PolSARPro software(ESA)and real ALOS PALSAR data over Amazon forest.The experiment results suggest that the proposed algorithm is more accurate than the classical three-stage method.