Model-based inversion algorithms based on bayesian compressive sensing

A model-based inversion algorithm is presented within the contrast source formulation of the inverse scattering problem. The inversion technique assumes that the scatterers at hand are limited to small areas within wide investigation domains or that they can be expressed in terms of very few unknown coefficients thanks to suitable representation bases. Accordingly, the associated inverse problem is solved through a Bayesian Compressive approach by enforcing a sparsity constraint in the retrieved contrast sources by means of a suitable prior. A preliminary numerical assessment is carried out to point out the features and the potentialities of the proposed technique.