SAR tomography from sparse samples

Three dimensional (3-D) Synthetic Aperture Radar (SAR) image formation provides the scene reflectivity estimation along azimuth, range and elevation co-ordinates. For 3-D image focusing multiple signals, acquired along different orbits, are required. The practical application of the focusing methods requires that non-uniformly spaced acquisition orbits have to be considered. In this paper we propose a technique exploiting the Compressive Sampling theory, and assuming that the image to be focused has a sparse representation along the elevation directions, which amounts to suppose that only few point-like scatterers with different elevation are present in the same range-azimuth resolution cell. Numerical results on simulated data show the good performance of the method.

1. Introduction

Synthetic Aperture Radar (SAR) systems provide two-dimensional (2-D) images of the observed scene [1], representing a projection of the 3-D scene scattering properties onto the 2-D range-azimuth plane.