Quantitative GPR Imaging via U-NET: Radargrams Versus Microwave Tomographic Inputs

Ground penetrating radar (GPR) imaging is mostly tackled by resorting to approximate linear inversion algorithms that provide only qualitative maps of the probed scene in terms of location and approximate geometry of the buried anomalies. Deep-learning (DL) techniques have recently been proposed to retrieve quantitative information as an alternative to classical non-linear inversion approaches. Indeed, deep neural networks can effectively learn to map the input data into spatial maps describing the electromagnetic (EM) properties of the targets. In this frame, the present article considers the popular U-NET topology for performing quantitative subsurface imaging. Two different training strategies differing for the type of input data are examined and compared. The first one assumes the radargram in the time domain as the network input; differently, in the second one, the network takes as input a microwave tomographic image of the subsurface scene. Numerical results based on full-wave synthetic data and some experimental tests are reported to assess and compare the reconstruction performance of both training schemes.