Anisotropic model-based SAR processing

Synthetic aperture radar (SAR) collections that integrate over a wide range of aspect angles hold the potentional for improved resolution and fosters improved scene interpretability and target detection. However, in practice it is difficult to realize the potential due to the anisotropic scattering of objects in the scene. The radar cross section (RCS) of most objects changes as a function of aspect angle. The isotropic assumption is tacitly made for most common image formation algorithms (IFA). For wide aspect scenarios one way to account for anistropy would be to employ a piecewise linear model. This paper focuses on such a model but it incorporates aspect and spatial magnitude filters in the image formation process. This is advantageous when prior knowledge is available regarding the desired targets’ RCS signature spatially and in aspect. The appropriate filters can be incorporated into the image formation processing so that specific targets are emphasized while other targets are suppressed. This is demonstrated on the Air Force Research Laboratory (AFRL) GOTCHA1 data set to demonstrate the utility of the proposed approach.