Fast High-Resolution 3D Radar Imaging Based on 3D FISTA

This paper presents a fast high-resolution three-dimensional (3D) imaging method for 3D imaging systems based on the 2D aperture and wideband waveforms. The proposed method realizes high-resolution imaging by exploiting the sparsity of the radar targets. Conventional sparse 3D imaging methods consume excessive time and memory due to converting the 3D problem into the ID problem. Herein, utilizing an existing 3D sparse representation, the proposed method directly solves the 3D imaging problem by a novel 3D fast iterative shrinkage-thresholding algorithm (3D FISTA) in the frequency-wavenumber domain. The memory load and computation complexity issues are solved successfully by avoiding transforming the 3D problem to the ID case. Experimental results demonstrate that our method provides accurate 3D image reconstruction at a low computation cost.