Towards Forest Structure Characteristics Retrieval from SAR Tomographic Profiles

Abstract

SAR Tomography has proven to be a unique tool for the retrieval of 3D structure information from forest scenarios: it can reveal different scattering mechanisms at different heights. However, the translation of these measurements into relevant forest structure information is not straightforward and research is still ongoing. In this direction, this paper suggest a framework for the estimation of forest structure from a SAR tomography scheme based on a low number of single pass coherences. Vertical reflectivity profiles are estimated by means of Compressive Sensing Imaging techniques. Two complementary descriptors are then suggested accounting for the simultaneous vertical and horizontal spatial variability of the scene. Their ability to reflect a characteristic structure behavior for the different types of forest considered is analyzed in simulated and real scenarios.