SAR tomography for point-like and volumetric scatterers using a regularised iterative adaptive approach

Abstract

ABSTRACTSynthetic aperture radar (SAR) tomography has been shown to be an effective and important tool for the monitoring of infrastructures and the investigation of forest vertical structure. However, the existing tomographic imaging algorithms, including spectral estimation and compressive sensing, can only be applied to SAR dataset over urban areas or over vegetated areas, while providing superresolution capability and working on single-look interferometric data. In this study, we propose a tomographic processing algorithm for SAR tomography based on Tikhonov regularisation and iterative adaptive approach (IAA), referred to as RIAA. As a nonparametric spatial spectral estimation method, RIAA allows the application to scenarios with point-like scatterers as well as with volumetric scatterers. The proposed method also provides superresolution capability and works on single-look SAR images. The performance of the proposed algorithm is verified by simulated data, TerraSAR-X strip-map and BioSAR 2007 P-band datasets over the urban and forested areas, respectively.