Inversion Algorithms and PS Detection in SAR Tomography. Case Study of Bucharest City

Abstract

Synthetic Aperture Radar (SAR) tomography can reconstruct the elevation profile of each pixel based on a set of co-registered complex images of a scene. Its main advantage over classical interferometric methods consists in the capability to improve the detection of single persistent scatterers as well as to enable the detection of multiple scatterers interfering within the same pixel. In this paper, three tomographic algorithms are compared and applied to a dataset of 32 images to generate the elevation map of dominant scatterers from a scene. Targets which present stable proprieties over time - Persistent Scatterers (PS) are then detected based on reflectivity functions reconstructed with Capon filtering.