多基线极化干涉SAR植被高度反演方法

摘要

This paper aims at the problems that the traditional single-baseline three stage inversion process is easily affected by noise factors, which will lead to a less precise inversion result.Two kinds of multi-baseline PolInSAR vegetation height inversion algorithms are adopted, which are the optimal method of geometric structure and the optimal method of vertical accuracy.Combined with the northern Sweden E-SAR P band data, the precision of the two algorithms for multi-baseline vegetation height inversion and their applied conditions in different vegetation areas is verified and obtained.The experimental result shows that the vegetation height inversion precision of the two methods has each increased by 44.05% and 22.11%, comparing with the traditional single-baseline algorithm.Besides, when the two methods works at the same time, it is better to adopt the optimal method of geometric structure when the vegetation height is more than 10m, while the optimal method of vertical accuracy is better when it is less than 10m.%针对传统单基线PolInSAR三阶段植被高度算法,反演易受到噪声因素影响导致反演结果精度较低等问题.文中采用几何结构最优法和高程精度最优法两种多基线PolInSAR植被高度反演方法,结合瑞典北部E-SAR P波段数据,验证得出两种多基线算法植被高度反演的精度以及其在不同植被高度覆盖区的适用条件.实验结果表明两种多基线方法植被高度反演精度相对传统单基线算法分别提高44.05%和22.11%,且当两种多基线方法同时反演植被高度时,植被高度大于10 m时采用几何结构最优法较好;植被高度小于10 m时采用高程精度最优法.