Radius estimation of subsurface cylindrical objects from ground-penetrating-radar data using full-waveform inversion

Ground-penetrating radar (GPR) crosshole tomography is widely applied to subsurface media images. However, the inadequacies of ray methods may limit the resolution of crosshole radar images, since the ray method is a type of high-frequency approximation. To solve this problem, the full waveform method is introduced for GPR inversion. However, full waveform inversion is computationally expensive. In this paper, we introduce a trained neural network that can be evaluated very quickly to replace a computationally intensive forward model. Additionally, the forward error of the trained neural network can be statistically analyzed. We demonstrate a methodology for a full waveform inversion of crosshole ground-penetrating radar data using the Markov chain Monte Carlo (MCMC) method. An accurate forward model based on Maxwell’s equations is replaced by a quickly trained neural network. This method achieves a high computation efficiency, which is four orders of magnitude faster than the accurate forward model. The inversion result of the synthetic waveform data shows a good performance of the trained neural network, which greatly improves the calculation efficiency.

show abstract

“…For the field data, it should be noted that the source wavelet needs to be estimated. Some methods can estimate the source wavelet [34,35].…”

Section: Experimental Validationmentioning

confidence: 99%

MCMC Method of Inverse Problems Using a Neural Network—Application in GPR Crosshole Full Waveform Inversion: A Numerical Simulation Study

et al. 2022

View full text Add to dashboard Cite

show abstract

“…However, it should be noted that the source wavelet needs to be estimated for the real data. Some estimation methods can solve this problem [38,39].…”

Section: Synthetic Examplesmentioning

confidence: 99%

Full-Waveform Inversion of Time-Lapse Crosshole GPR Data Using Markov Chain Monte Carlo Method

et al. 2021

View full text Add to dashboard Cite

Crosshole ground-penetrating radar (GPR) is an important tool for a wide range of geoscientific and engineering investigations, and the Markov chain Monte Carlo (MCMC) method is a heuristic global optimization method that can be used to solve the inversion problem. In this paper, we use time-lapse GPR full-waveform data to invert the dielectric permittivity. An inversion based on the MCMC method does not rely on an accurate initial model and can introduce any complex prior information. Time-lapse ground-penetrating radar has great potential to monitor the properties of a subsurface. For the time-lapse inversion, we used the double difference method to invert the time-lapse target area accurately and full-waveform data. We propose a local sampling strategy taking advantage of the a priori information in the Monte Carlo method, which can sample only the target area with a sequential Gibbs sampler. This method reduces the calculation and improves the inversion accuracy of the target area. We have provided inversion results of the synthetic time-lapse waveform data that show that the proposed method significantly improves accuracy in the target area.

show abstract

“…), GPR can be used to detect both metallic and non-metallic targets. In addition to the wide range of GPR applications listed in [2], estimating the depth and radius of buried cylindrical pipes has become an important task-for instance, in concrete rebar investigation and underground utility network localization [3]. However, regardless of the information acquired using GPR, each application requires suitable processing techniques in order to interpret GPR data and for decision-making purposes.…”

Section: Introductionmentioning

confidence: 99%

“…However, regardless of the information acquired using GPR, each application requires suitable processing techniques in order to interpret GPR data and for decision-making purposes. Within the scope of buried utility pipes, since the 3D localization of underground utility pipes has become mandatory to avoid accidents during excavation, the estimation of depth and radius has been widely studied, as demonstrated in the literature, using the following: the ray-based method [4], full-wave inversion (FWI) [3], Hough transforms [5] and machine learning techniques [6]. Recently, Liu et al [3] used ray-based and FWI approaches to develop a novel method to estimate radius, depth and relative permittivity of utility pipes.…”

Section: Introductionmentioning

confidence: 99%

A Preliminary Numerical Study to Compare the Physical Method and Machine Learning Methods Applied to GPR Data for Underground Utility Network Characterization

et al. 2022

View full text Add to dashboard Cite

In the field of geophysics and civil engineering applications, ground penetrating radar (GPR) technology has become one of the emerging non-destructive testing (NDT) methods thanks to its ability to perform tests without damaging structures. However, NDT applications, such as concrete rebar assessments, utility network surveys or the precise localization of embedded cylindrical pipes still remain challenging. The inversion of geometric parameters, such as depth and radius of embedded cylindrical pipes, as well as the dielectric parameters of its surrounding material, is of great importance for preventive measures and quality control. Furthermore, the precise localization is mandatory for critical underground utility networks, such as gas, power and water lines. In this context, innovative signal processing techniques associated with GPR are capable of performing physical and geometric characterization tasks. This paper evaluates the performance of a supervised machine learning and ray-based methods on GPR data. Support vector machines (SVM) classification, support vector machine regression (SVR) and ray-based methods are all used to correlate information about the radius and depth of embedded pipes with the velocity of stratified media in various numerical configurations. The approach is based on the hyperbola trace emerging in a set of B-scans, given that the shape of the hyperbola varies greatly with pipe depth and radius as well as with velocity of the medium. According to the ray-based method, an inversion of the wave velocity and pipe radius is performed by applying an appropriate nonlinear least mean squares inversion technique. Feature selection within machine learning models is also implemented on the information chosen from observed hyperbola travel times. Simulated data are obtained by means of the finite-difference time-domain (FDTD) method with the 2D numerical tool GprMax. The study is carried out on mono-static, ground-coupled GPR datasets. The preliminary study showed that the proposed machine learning methods outperforms the ray-based method for estimating radius, depth and velocity. SVR, for instance, calculates depth and radius values with mean absolute relative errors of 0.39% and 6.3%, respectively, with regard to the ground truth. A parametric comparison of the aforementioned methodologies is also included in the performance analysis in terms of relative error.

show abstract

Radius estimation of subsurface cylindrical objects from ground-penetrating-radar data using full-waveform inversion

Cited by 28 publications

References 46 publications

MCMC Method of Inverse Problems Using a Neural Network—Application in GPR Crosshole Full Waveform Inversion: A Numerical Simulation Study

MCMC Method of Inverse Problems Using a Neural Network—Application in GPR Crosshole Full Waveform Inversion: A Numerical Simulation Study

Full-Waveform Inversion of Time-Lapse Crosshole GPR Data Using Markov Chain Monte Carlo Method

A Preliminary Numerical Study to Compare the Physical Method and Machine Learning Methods Applied to GPR Data for Underground Utility Network Characterization

Contact Info

Product

Resources

About