Multi-Resolution Bayesian Compressive Sensing for Sparse Target Inversion

SALUCCI, Marco and POLI, Lorenzo and ZARDI, Francesco and TOSI, Luca and LUSA, Samantha and MASSA, Andrea (2025) Multi-Resolution Bayesian Compressive Sensing for Sparse Target Inversion. Technical Report. ELEDIA Research Center - University of Trento.

[img] Text
Multi-Resolution_Bayesian_Compressive_Sensing_for_Sparse_Target_Inversion.pdf

Download (1MB)

Abstract

Non-Born scatterer retrieval is addressed within the contrast source inversion (CSI) framework using a novel multi-step inverse scattering approach. The unknown contrast sources are represented in a multi-resolution (MR) framework and iteratively reconstructed using a Bayesian compressive sensing (BCS) approach, which employs a constrained relevance vector machine solver to promote sparsity. To assess the reliability and robustness of the proposed MR-BCS-CSI method, representative inversions from both synthetic and experimental data are presented. Additionally, comparisons with recent state-of-the-art alternatives are provided.

Item Type: Monograph (Technical Report)
Uncontrolled Keywords: microwave imaging (MI), contrast source inversion (CSI), Bayesian compressive sensing (BCS), inverse scattering (IS)
Subjects: A Areas > A WC Next Generation Wireless Communications
M Methodologies > M CS Compressive Sensing
URI: http://www.eledia.org/students-reports/id/eprint/911

Actions (login required)

View Item View Item