A linear inverse scattering algorithm for realistic GPR applications

The paper deals with a microwave-tomography-based solution algorithm tailored for use with GPR data-processing applications. The algorithm tackles an inverse scattering problem in the frequency domain through the use of a linear model of the electromagnetic scattering, based on the Born approximation. In particular, we evaluate the reconstruction capabilities of the linear inversion algorithm in terms of the retrievable spatial variations of the unknown contrast function, whilst considering the problem of choosing an optimal frequency measurement step, theoretically, using diffraction tomography arguments. A numerical analysis of the technique is performed by means of the singular-value decomposition tool, which allows us to extend the theoretical results to more realistic cases involving lossy soils. Finally, we present a series of reconstructions, obtained using synthetic and experimental data, which show the performance of the method under realistic conditions.