Bayesian compressive sensing for synthetic-aperture radar tomography imaging

Abstract

To achieve high-resolution three-dimensional images, a number of imaging methods based on compressive sensing (CS) have been suggested in the recent years for synthetic-aperture radar (SAR) tomography. However, the CS-based methods are sensitive to noise. In this work, we develop a new Bayesian compressive sensing (BCS) imaging method for the SAR tomography. In the framework of BCS, a 'sparseness' prior distribution of the imaging scene and an additive noise are properly considered in the imaging process. As a consequence, the BCS-based method under the conditions of low noise levels can provide a better performance than the common norm-based CS methods. The results obtained via simulations of our SAR-tomography imaging method confirm its advantages.