Robust Compressed Sensing recovery for detecting two-dimensional scatters

In this paper, a robust compressive sensing (CS) recovery method for detecting and imaging two-dimensional scatters is proposed. The proposed method reformulates the inverse scatter problem into a Bounded-Constraint Quadratic Program (BCQP) which can be solved efficiently by gradient projection. Compared with the state-of-the art Bayesian Compressed Sensing (BCS) inversion method, simulation verified that the proposed method preforms more robust and effective in low SNR, mutual effect and non-sparse scenarios.