3D SAR Iterative Imaging Algorithm Via Joint Low-Rank and Sparsity

Abstract

SAR sparse imaging based on the compressed sensing (CS) method has shown great potential for high-resolution imaging, reducing sampling rate, and denoising. However, conventional sparse imaging algorithms always encounter complex matrix calculation and enormous iterations. To solve these problems, in this article, we propose a novel low-rank 3D SAR sparse imaging algorithm, i.e., Iterative Imaging Algorithm via Joint Low-rank and Sparsity (LSIIA), which is structured based on the radar perception model to achieve the accurate sparse reconstruction at a low sampling rate. To improve the computing efficiency, the frequency-domain imaging operator is introduced into the algorithm. The low-rank condition constrains the redundancy of data, and $$\ell_{1}$$ norm characterizes the sparsity of data. Joint the low-rank condition and $$\ell_{1}$$ regularization, and the Half Quadratic Splitting (HQS) method is applied to simplify the iteration. LSIIA can produce high-quality 3D SAR images under a low sampling rate. The simulation and experiments show the superiority and effectiveness of the proposed algorithm.