A fast parallel primal-dual based compressed sensing method for 2-D near-field SAR image reconstruction

摘要

In this paper, a fast compressed sensing (CS) method for 2-D near-field synthetic aperture radar (SAR) image reconstruction from under-sampled measurements is proposed. A parallel primal-dual process as core steps is applied in this method with minimization of the multiple penalized least squares model which is the combination of the least squares data fidelity term, ℓ₁-norm, total variation (TV) operator and an indicator function for bounding. Experimental results on real data demonstrate the effectiveness and efficiency of the proposed method.