Comparison of Short-Range SAR Imaging Algorithms for the Detection of Landmines using Numerical Simulations

This paper deals with the compensation of imaging errors in the field of landmine detection with handheld devices, caused by the varying propagation behavior of electromagnetic waves in the surrounding medium. For this purpose, an analytical model will be introduced, which yields a description of the refraction behavior for planar surfaces and therefore, allows to determine the traveled distance of the wave. In addition, it will be presented, how this solution can be used in the field of SAR imaging for ground penetrating radar applications in the case of non-planar surfaces. Finally, the results will be tested on data sets, suitable to train a deep neural network. Those data sets were created with a numerical simulator, which calculates the received signal for a given ground scene using finite differences.