A Bayesian-Compressive-Sampling-Based Inversion for Imaging Sparse Scatterers

Author

Oliveri, G. ; Rocca, Paolo ; Massa, A.

Author_Institution

Dept. of Inf. Eng. & Comput. Sci., ELEDIA Res. Group, Univ. of Trento, Trento, Italy

Volume

49

Issue

10

fYear

2011

Firstpage

3993

Lastpage

4006

Abstract

In this paper, a new approach based on the Bayesian compressive sampling (BCS ) and within the contrast source formulation of an inverse scattering problem is proposed for imaging sparse scatterers. By enforcing a probabilistic hierarchical prior as a sparsity regularization constraint, the problem is solved by means of a fast relevance vector machine. The effectiveness and robustness of the BCS-based approach are assessed through a set of numerical experiments concerned with various scatterer configurations and different noisy conditions.

Keywords

Bayes methods; geophysical techniques; imaging; Bayesian-compressive-sampling-based inversion; contrast source formulation; fast relevance vector machine; inverse scattering problem; microwave imaging; numerical experiments; probabilistic hierarchical prior; scatterer configurations; sparse scatterers; sparsity regularization constraint; Bayesian methods; Image reconstruction; Inverse problems; Microwave imaging; Signal to noise ratio; Bayesian compressive sampling (BCS); contrast source formulation; inverse scattering; microwave imaging; relevance vector machine (RVM);

fLanguage

English

Journal_Title

Geoscience and Remote Sensing, IEEE Transactions on

Publisher

ieee

ISSN

0196-2892

Type

jour

DOI

10.1109/TGRS.2011.2128329

Filename

5766031