High-resolution radar imaging for non-sparse scatterers by the combination of non-convex regularization and total variation

Most existing compressed sensing (CS) based radar imaging methods are based on the assumption that the targets are sparse enough, while in practice the targets are often spatially extended, which would degrade their inversion performances severely. In fact, the concentrations of the corresponding strong scatterers always form certain regions in high-resolution radar. Therefore, there still exist dependence and redundancy needed to exploit. In this paper, by utilizing the sparsity and continuity property of the targets, we propose a novel CS-based method for the radar imaging of two dimensional non-sparse scatterers, which is combining non-convex regularization and total variation constraint. Experimental results verify the effectiveness of the proposed method.