SAR tomography of forested areas: An APES-based virtual beamforming approach

Abstract

The main purpose of synthetic aperture radar tomography (T-SAR) is to retrieve the vertical distribution of the backscattered power within a range-azimuth resolution cell, allowing three dimensional (3-D) imaging of volumetric targets. The achievable resolution of conventional spectral estimators, used to tackle with the T-SAR vertical distribution estimation problem, highly depends on the span of the tomographic aperture and on the number of acquisitions, reason why, adaptive spectral estimators were introduced in order to ease such limitations. This paper proposes a novel amplitude and phase estimation (APES) -based virtual beamforming approach aimed at resolution enhanced SAR tomography of forested areas, which functions properly using only a few passes and arbitrary constellation geometries, and with imaging capabilities comparable to those gotten from super-resolution compressed sensing spectral estimators with the sum of Kronecker products decomposition technique as a pre-processing step.