Advanced Signal Processing for Ground-Penetrating Radar

Abstract

Ground-penetrating radar (GPR) is a high-frequency electromagnetic detection device used for land mine detection and other relevant applications such as identification of cement structures, underground metal pieces, and many more fields. This paper addresses the problem of analyzing and improving GPR performance in detecting and characterizing objects in the background of clutter and noise. Significant amount of works in this area have been attempted recently. But, while eliminating direct waves, the traditional time domain method introduces false signals and conflicts and does not provide good random noise elimination. The problem associated with this work is the extraction of the required target signal buried/subdued under noise and clutter whose characteristics keep on changing with the types of soils and external environment. This paper focuses on the enhancement of target information in the background of clutter and noise, by using the simulated data for the comparison and evaluation of different algorithms used for GPR signal processing, against measured data in reduction of clutter and unwanted noise.