Ground Penetrating Radar Imaging for Buried Cavities in a Dispersive Medium: Profile Reconstruction Using a Modified Hough Transform Approach and a Time-Frequency Analysis

Abstract

Ground Penetration Radar is a non-destructive tool to explore the underground soil. The modeling is a very important process in order to interpret the GPR profiles. Solving the equations of Maxwell at this level, for a dispersive medium, is modeled by the finite difference method (FDTD) with UPML (uniaxial perfectly matched layers) boundary conditions, because it's an easier and an efficient numerical technique. In this paper, we propose an easier approach to estimate physical parameters of buried targets, of ground penetration radar images (B-scans Profiles), such as voids and small cavities. Traditionally, we use the hough transform to detect the hyperbolas caused by targets reflections, but this one is such difficult in implementation, because in this case we deal with a three-dimensional parametric space. The performance of target recognition depends on the preprocessing phase. A filtering schema based on singular value decomposition (SVD) is proposed to eliminate the direct wave or the clutter, as a first step. Second, we use wavelet decomposition for denoising the B-scan image, a special reconstruction schema is proposed, a B-scan fully reconstructed and one reconstructed ignoring the horizontal component, and a cross-correlation calculation between the two is introduced, in order to enhance the targets signal (hyperbolas). Hyperbola’s apexes are detected using a corner detection approach. Besides, the nature of the dielectric in the cavity, is estimated using a time-frequency analysis, due to the fact that the reflections are not visible neither in the A-scan traces or their spectrum.
Copyright © 2015 Praise Worthy Prize - All rights reserved.