A Non-Uniform Sub-Aperture Fast Back-Projection Algorithm for Spaceborne SAR Terrain Matching Curved Imaging

Conventional synthetic aperture radar (SAR) generates the imaging swath parallel to the satellite trajectory without considering about the complex orientation of scene, leading to the inefficient multiple-swaths observation of the long curved terrains, like seismic faults, railways, coastlines, etc.. Nowadays, spaceborne SAR terrain matching (TM) curved imaging is proposed to solve this problem by generating a flexible curved swath matching with the long curved terrain, improving the observing efficiency successfully. However, there is currently no suitable imaging algorithm that fully considers the complex parameters in the spaceborne SAR TM curved imaging. In this paper, we proposed a new fast back-projection algorithm for the spaceborne SAR TM curved imaging. The main technical contributions are: First, a nonuniform sub-aperture dividing method is proposed to adapt to the spatial variance of wavenumber spectrum. Second, a non-orthogonal grid meshing method is proposed to fit the complex orientation of the TM curved swath. Third, a modified spectrum compression method is proposed to accelerate processing. A spaceborne SAR TM curved imaging mission of observing the main urban area of Wuhan in China is performed with the Chinese LJ2-01 satellite. Real data results verify the effectiveness of the proposed algorithm.