基于多径模型的建筑区SAR三维成像中非视距目标重定位方法

Non-line-of-sight Target Relocation by Multipath Model in SAR 3D Urban Area Imaging

随着合成孔径雷达(SAR)系统小型化技术、SAR三维成像技术的发展,当前已可利用无人机载小型化阵列干涉SAR实现城区三维成像,在城市测绘、复杂环境重建等领域具有重要应用前景。然而,SAR进行城市场景三维成像时,回波中存在多径信号,会导致成像结果解译困难,却也给非视距区域隐蔽目标的发现提供了重要手段。为此,该文针对低空无人机载阵列干涉SAR建筑区三维成像中的非视距目标进行了研究,建立了非视距目标在低空阵列干涉三维成像下的多径模型,给出了城市峡谷区域利用多径扩大可视范围的计算方法,并基于建筑平面拟合提出了非视距目标重定位方法。无人机载阵列干涉SAR的仿真和实际数据处理验证表明,所提出的方法可以对城市峡谷非视距目标进行有效的三维成像和重定位,重定位误差小于0.5 m,实现了对非视距区域信息的获取。

The advancement in the miniaturization technology of Synthetic Aperture Radar(SAR)systems and SAR three-dimensional(3D)imaging has enabled the 3D imaging of urban areas through Unmanned Aerial Vehicle(UAV)-borne array Interferometric SAR(array-InSAR),offering significant utility in urban cartography,complex environment reconstruction,and related domains.Despite the challenges posed by multipath signals in urban scene imaging,these signals serve as a crucial asset for imaging hidden targets in Non-Line-of-Sight(NLOS)areas.Hence,this paper studies NLOS targets in UAV-borne array-InSAR 3D imaging at low altitudes and establishes a multipath model for 3D imaging at low altitudes.Then,a calculation method is proposed for obtaining the multipath reachable range in urban canyon areas based on building plane fitting.Finally,a relocation method for NLOS targets is presented.The simulation and real data experiments of UAV-borne array-InSAR show that the proposed method can effectively obtain 3D images and relocate NLOS targets in urban canyon areas,with errors typically below 0.5 m,which realizes the acquisition of hidden NLOS region information.