The Experimental Research On Ground Penetrating Radar Imaging

Ground Penetrating Radar (GPR) may be used to explore the electrical structuredistribution of near-surface medium by reflection of high-frequency electromagnetic waves.As a high-resolution geophysical method with some characteristics such as high detectionaccuracy, flexible operation, and name a few, GPR has widely been applied to a variety offields, for example, the engineering geological exploration, engineering quality detection,archaeological probing, and so on. The depth and accuracy of GPR exploration is mainlydecided by conductivity, dielectric constant, the antenna center frequency, etc., and ismainly affected by the space-time sampling rate, the distance of transmitting and receivingantenna, the interference from detection site, and so forth. How to accurately select theappropriate system parameters and modes of observation, and how to deal with theinterference factors of exploration site, as well as an accurate estimate electrical parametersof the exploration area are the problems to be solved in GPR engineering practice.This paper analyzes and summarizes the theoretical resolution limit of GPR in sometypical subsurface medium based on propagation rules of electromagnetic wave and workprinciples of GPR. Using a tool of forward modeling, we give a detailed analysis of thefactors that affect the depth and resolution of GPR, and discuss the impact of the electricalparameters, system parameters, and the distance of transmitting and receiving antenna onthe accuracy of GPR imaging. At the same time, we research and provide differentdenoising methods for field data. Moreover, we study the discriminative ability of GPRimaging by the calibration with high-definition drilling video.The main productions of this paper are described as follows.(1) Based on numerical simulations, this paper is the first to analyze systematically the impact of the electrical parameters and system parameters on GPR imaging.(2) This paper shows a systematic analysis of the FIR filter, wavelet transform andprincipal component analysis methods to abate the effect of random noise and coherentnoise, and presents a conclusion that the FIR band-pass filter and wavelet transformhave the better ability to suppress random interference, while the principal componentanalysis filter can obtain the preferable results to eliminate coherent noise.(3) This paper presents the first experiment to study the discriminative ability of GPRimaging of fine geological structure with centimeter-level by the calibration withhigh-definition drilling video, and verifies that GPR can discriminate the geologicalstructure with centimeter-level in operating frequency with MHz level. Moreover, theexperiment also shows that the dispersion of GPR is serious and has a significantimpact on the interpretation of results.