基于去相干敏感因子的多基线PolInSAR植被高度反演方法

Multi-baseline PolInSAR Vegetation Height Inversion Method Based on Decorrelation Sensitivity Factor

极化干涉合成孔径雷达(PolInSAR)技术通过建立相关质量评价指标筛选最优干涉几何,从而实现多基线植被高度反演。但目前使用的相干性特性、测高精度等质量指标并未关联与干涉几何直接相关的垂直波数,容易降低干涉几何选取的稳健性。本文提出一种基于去相干敏感因子的多基线PolInSAR植被高度反演方法,利用敏感因子描述干涉几何与植被高度的匹配程度,从多个单基线固定消光方法中确定最佳植被高度反演结果。使用欧洲航天局AfriSAR 2016项目的P波段合成孔径雷达(SAR)数据进行实验验证。实验结果表明:以激光雷达(Light Detection and Ranging,LiDAR)获取的植被高度作为参考,本文多基线反演算法的均方根误差为6.19 m,比其他两种传统方法精度分别提高了约14.14%和13.55%。

Polarimetric interferometric synthetic aperture radar(PolInSAR)technology carries out multi-baseline vegetation height inversion by using relevant quality evaluation index to select the optimal interferometric geometry.However,the currently used quality indices such as coherence separation product and height accuracy are not associated with the vertical wavenumber directly related to the interferometric geometry,which may reduce the robustness of the interferometric geometry selection.This paper proposes a multi-baseline PolInSAR vegetation height inversion method based on decorrelation sensitivity factor,which can describe the matching degree between interferometric geometry and vegetation height.The optimal vegetation height is determined from multiple single-baseline inversion results using fixed extinction coefficient method.This paper uses P-band synthetic aperture radar(SAR)data collected by the European Space Agency AfriSAR 2016 campaign for experimental validation.The experimental results show that with the light detection and ranging(LiDAR)vegetation height as a reference,the root mean squared error(RMSE)of the proposed multi-baseline inversion algorithm is 6.19 m,which is about 14.14%and 13.55%higher than those of the other two traditional methods.