Performance of free-space tomographic imaging approximation for shallow-buried target detection

Forward-looking ground penetrating radar (FL-GPR) is an emerging modality that permits standoff sensing of targets buried at shallow depths in the ground. Most FL-GPR imagery is obtained using free-space approximation, neglecting the presence of the air-to-ground interface and assuming the propagation as occurring in a homogeneous dielectric medium. In this paper, we compare the performance of the approximate free-space tomographic imaging with that of a tomographic algorithm which accounts for the presence of the actual halfspace geometry. The half-space approach implements the spectral representation of the dyadic Green's function. Using numerical electromagnetic FL-GPR data, we investigate the impact of the free-space approximation on the image quality as well on the image-domain statistics of the targets and rough surface clutter.