基于压缩感知的星载SAR差分层析高分辨成像

High Resolution Imaging for Differential SAR Tomography Based on Compressive Sensing

城区建筑的四维成像是合成孔径雷达差分层析的重要应用领域之一。此种应用背景下,如何利用空间-时间二维平面内稀疏分布的观测数据,在保持方位向-距离向分辨率的同时实现高程向-形变速率向的高分辨成像是合成孔径雷达差分层析面临的难点问题。压缩感知为该问题提供了有效的解决方案。首先,对压缩感知在差分层析中应用的可行性进行了理论分析。随后,利用仿真实验分析了基于压缩感知的星载合成孔径雷达差分层析成像的高程向和形变速率向的分辨能力。最后,使用Envisat-ASAR数据进行了实测数据处理实验,取得了理想的成像结果。基于压缩感知的星载合成孔径雷达差分层析高分辨成像方法的有效性和实用性得到了验证。

Four-dimensional reconstruction of urban areas and man-made infrastructure is one of the most important applications of differential SAR（Synthetic Aperture Radar） tomography.In this situation,the urgent problem of differential SAR tomography is how to achieve satisfying elevation-velocity resolution using sparse samples of space-time plane and preserve azimuth-range resolution simultaneously.Compressive sensing is a perfect theory to resolve this urgent problem.First,we prove the applicability of compressive sensing for differential SAR tomography theoretically.Then,the resolution ability of differential SAR tomography based on compressive sensing is discussed using simulation experiment with the real system parameters.Finally,we apply this imaging method to the real data of Envisat-ASAR.Both the theoretical analysis and satisfying results of Envisat-ASAR data processing confirmed the applicability of this differential SAR tomograhic imaging method.