Simulation study of the effect of certain types of noises on the quality of ground penetrating radar (GPR) data

Abstract

Ground penetrating radar (GPR) is a valuable measuring tool and its fields of use are multiple (civil engineering, space exploration, mine clearance, etc.). During the inspection of the terrain, by the GPR radar, 'noises' can appear on the image and mask the desired information. These parasitic signals (noise) can come from internal sources (such as direct coupling between antennas, system rings) or external sources generated outside the system (such as: reinforced concrete bridges, building pillars). This work presents a simulation study of the influence of two types of noise on the quality of GPR data. These noises are the white Gaussian noise which resembles the background noise during GPR measurements and the noise caused by the direct coupling between the GPR antennas. The influence of the latter was studied with different distances between the antennas, using the operating mode of the common GPR receiver. The results obtained showed that Gaussian noise causes a complete overlap of the GPR images, while the direct coupling between the antennas causes the appearance of a direct wave, of variable depth depending on the distance between the antennas. This led to the masking of certain parts of the hyperbolas showing the presence of the targets