SAR correlation imaging and anisotropic scattering

Abstract

In this paper we investigate the ability of correlation syntheticaperture radar (SAR) imaging to reconstruct isotropic and anisotropic scatterers. SAR correlation imaging was suggested by the author previously in [34]. Correlation imaging algorithms produce an image of a second-order quantity describing an object an interest, for example, the reflectivity function squared. In the previous work [34] it was argued that the effects of volume scattering clutter on the image can be minimized by choosing which pairs of collected data to correlate prior to applying a backprojection-type reconstruction algorithm. This choice of pairs for the correlation process is determined by what is known as the memory effect of scattering of waves by random scatterers [42, 43, 40, 41, 7, 14]. It is the goal of this current work to determine the different imaging outcomes for an isotropic or point scatterer versus an anisotropic or dipole scatterer. In addition we aim to determine if removing contributions to the image due to the memory effect is necessary for diminishing the contributions of anisotropic or clutter scatterers to the scene of interest. Finally we extend the analysis of [34] to the polarimetric SAR case to determine whether the additional data provided by this modality contributes to decreasing the effects of clutter on the SAR image.

Compound truncated Poisson normal distribution: Mathematical properties and Moment estimation

“…Now, we employ the CTPN to real data extracted from SAR images. SAR system is one of the most successful remote sensing techniques [17]. This fact can be justified by its capability of operating under all weather conditions and of producing high spatial-resolution images.…”