%0 Report
%9 Technical Report
%A Oliveri, G.
%A Salucci, M.
%A Anselmi, N.
%D 2018
%F elediasc12:785
%K Born approximation (BA), Bayesian Compressive Sensing (BCS), Inverse Scattering (IS), Matrix Completion (MC)
%T A Matrix Completion Approach to Image Sparse Targets under the First‐Order Born Approximation
%U http://www.eledia.org/students-reports/785/
%X In this work, a novel approach to solve the microwave inverse scattering problem under the first-order Born approximation to image sparse and weak targets is addressed. Towards this end, a single-task Bayesian compressive sensing (ST-BCS) solver is exploited to retrieve a preliminary guess of the unknown contrast distribution within an inaccessible domain. Then, a filtering process is used to filter out the "less reliable" contrast coefficients from the BCS guess. Finally, a customized matrix completion (MC) procedure is adopted in order to complete the retrieved images and achieve significant accuracy improvements under high levels of noise on processed data.   In order to validate the effectiveness of the proposed methodology some representative numerical benchmarks are presented considering different targets and noise levels.