採用無人機群執行高解析度 L 波段層析合成孔徑雷達成像以檢測森林冠層介電常數異常

Abstract

本論文提出了嚴格的層析合成孔徑雷達 (TomoSAR) 成像程序來獲取局部感興趣區域森林的高解析度 L 波段影像。推導出一聚焦函數 (Focusing function) 以將散射訊號與森林冠層的反射率函數相關聯，從而無需借助校準。使用森林體素模型來模擬不同的樹種，並使用 Maxwell-Garnett mixing formula 對介電常數進行建模。在模擬中，在三種信噪比下的兩種森林場景中應用了五種不同的反演方法，以驗證所提出程序的有效性。樹木的介電常數曲線可用來監測森林的含水量。結果顯示使用一群無人機 (UAV) 在特定區域進行 TomoSAR 成像以找出潛在的野火危險點是可行的。

A rigorous TomoSAR imaging procedure is proposed to acquire high-resolution L-band images of a forest in a local area of interest. A focusing function is derived to relate the backscattered signals to the reflectivity function of the forest canopies without resorting to calibration. A forest voxel model is compiled to simulate different tree species, with the dielectric constant modeled with the Maxwell-Garnett mixing formula. Five different inverse methods are applied on two forest scenarios under three signal-to-noise ratios in the simulations to validate the efficacy of the proposed procedure. The dielectric-constant profile of trees can be used to monitor the moisture content of the forest. The use of a swarm of unmanned aerial vehicles (UAVs) is feasible to carry out TomoSAR imaging over a specific area to pinpoint potential spots of wildfire hazards.