Interpretation Of Ground Penetrating Radar Data Research Articles

Soil water content estimation by using ground penetrating radar data full waveform inversion with grey wolf optimizer algorithm

AbstractSoil water content (SWC) estimation is important for many areas including hydrology, agriculture, soil science, and environmental science. Ground penetrating radar (GPR) is a promising geophysical method for SWC estimation. However, at present, most of the studies are based on partial information of GPR, like travel time or amplitude information, to invert the SWC. Full waveform inversion (FWI) can use the information of the entire waveform, which can improve the accuracy of parameter estimation. This study proposes a novel SWC estimation scheme by using the FWI of GPR, optimized by the grey wolf optimizer (GWO) algorithm. The proposed scheme includes a petrophysical relationship to link the SWC with the relative dielectric permittivity, 1D GPR forward modeling, and a GWO optimization algorithm. First, numerical modeling was carried out, and the proposed scheme was applied to both noise‐free and noisy data to verify its applicability. Then, the proposed method was applied to data collected from a field experimental site. These results, derived from both synthetic and real datasets, show that the proposed inversion scheme resulted in a good match between the observed and calculated GPR data. In the numerical modeling, it was observed that the SWC could be inverted accurately, even when noise was present in the data. These demonstrate that the GWO method can be applied for the quantitative interpretation of GPR data. The proposed scheme shows potential for SWC estimation by using GPR full waveform data.

3D ground-penetrating radar data analysis and interpretation using attributes based on the gradient structure tensor

In near-surface geophysics, ground-penetrating radar (GPR) surveys are routinely used in a variety of applications including those from archaeology, civil engineering, hydrology, and soil science. Thanks to recent technical developments in GPR instrumentation and antenna design, 3D surveys comprising several hundred thousand traces can be performed daily. Especially in complex environments such as sedimentary systems, analyzing and interpreting the resulting GPR volumes is a time-consuming and laborious task that is still largely performed manually. In the past few decades, several data attributes have been developed to guide and improve such tasks and assure a higher degree of reproducibility in the resulting interpretations. Many of these attributes have been developed in image processing or computer vision and are routinely used, for example, in reflection seismic data interpretation. Especially in sedimentary systems, variations in the subsurface are accompanied by variations of GPR reflections in terms of the amplitudes, continuity, and geometry in view of the dip angle and direction. A promising tool to analyze such structural features is known as the gradient structure tensor (GST). To date, the application of the GST approach has been limited to a few 2D GPR examples. Thus, we take the basic idea of GST analysis and introduce and evaluate the corresponding attributes to analyze 3D GPR data. We apply our GST approach to one synthetic and two field data sets imaging diverse sedimentary structures. Our results demonstrate that our set of GST-based attributes can be efficiently computed in three dimensions and that these attributes represent versatile measures to address different typical interpretation tasks and, thus, help for an efficient, reproducible, and more objective interpretation of 3D GPR data.

CycleGAN-Based Data Augmentation for Subgrade Disease Detection in GPR Images with YOLOv5

Vehicle-mounted ground-penetrating radar (GPR) technology is an effective means of detecting railway subgrade diseases. However, existing methods of GPR data interpretation largely rely on manual identification, which is not only inefficient but also highly subjective. This paper proposes a semi-supervised deep learning method to identify railway subgrade diseases. This method addresses the sample imbalance problem in the defect dataset by utilizing a data augmentation method based on a generative adversarial network model. An initial network model for disease identification is obtained by training the YOLOv5 network with a small number of existing samples. The intelligently extended samples are then labeled to achieve a balance in the disease samples. The network is trained to improve the recognition accuracy of the intelligent model using a more complete dataset. The experimental results show that the accuracy of the proposed method can reach up to 94.53%, which is 23.85% higher than that of the supervised learning model without an extended dataset. This has strong industrial application value for railway subgrade disease detection as the potential learning ability of the model can be explored to a greater extent, thereby improving the recognition accuracy of subgrade diseases.

Application of Variational Mode decomposition in GPR Detection of Oil Contaminated Soil

In this article, Ground penetrating radar (GPR) is used to investigate the contaminated soil by oil pollutants in oilfield operation area. Time frequency analysis method is applied to GPR data to reveal the changes of GPR signal frequency components with time variation, which is very important for data interpretation. We use variational mode decomposition (VMD) for time frequency analysis. It can decompose radar signal into several quasi-orthogonal intrinsic mode functions (IMFs) with limited bandwidth, and has strong robustness and local decomposition ability. We apply the VMD on field data acquired in the area with waste liquid contaminated soil in oilfield. The decomposed IMFs from GPR data shows that the signals from the contaminated soil is characterized by the strong energy and low frequency. Then the instantaneous attribute analysis of the IMF is carried out to delineate the scope and thickness of contaminated soil. The experimental results show that the limited bandwidth IMF with specific sparsity obtained by VMD method is more conducive to the processing and interpretation of GPR data.

A framework for GPR‐based water leakage detection by integrating hydromechanical modelling into electromagnetic modelling

AbstractTimely and accurate detection of water pipe leakage is critical to preventing the loss of freshwater and predicting potential hazards induced by the change in underground water conditions, thereby developing mitigation strategies to improve the resilience of pipeline infrastructure. Ground‐penetrating radar (GPR) has been widely applied to investigating ground conditions and detecting pipe leakage. However, due to uncertainties of complex underground environments and time‐lapse change, a proper interpretation of GPR data has been a challenging task. This paper aims to leverage hydromechanical (HM) modelling to predict electromagnetic (EM) responses of water leakage detection in diverse leakage cases. A high‐fidelity 3D digital model of an actual pipeline network, hosting pipes with various sizes and materials, was reconstructed to represent the complex geometry and various mediums. The interoperability between the digital model and the numerical models utilized in HM and EM simulations was improved to better capture the irregular pipelines. Based on Kriging interpolation and the volumetric complex refractive index model, a linking technique was employed to replicate material heterogeneity caused by water intrusion. Thus, a framework was developed to accommodate the interoperability among digital modelling, HM modelling and finite‐difference time‐domain forward modelling. Moreover, sensitivity studies were conducted to evaluate the influences of different time stages, leak positions and pipe types on GPR responses. In GPR B‐scans, the presence of hyperbolic motion and horizontal reflections serve as indicators to estimate the location and scale of water leakage. Specifically, a downward‐shifting hyperbola indicates that the pipeline is submerged by leaked water, whereas the emergence of horizontal reflection is linked to the wetting front of saturated areas. The developed framework can be expanded for complicated applications, such as unknown locations and unforeseen failure modes of pipelines. It will increase the efficiency and accuracy of traditional interpretations of GPR‐based water leakage detection and thus enable automated interpretations by data‐driven methods.

Reflection characteristics of typical road defects in 3D GPR images for collapse mitigation

Road collapses pose dangerous threats to traffic safety in urban cities. Typically, these collapses result from under-road defects such as cavities, loose soil, and voids. With the development of multi-channel 3D ground penetrating radar (GPR) systems, GPR has become a routine method for inspecting the subsurface conditions of roads. However, the reflection characteristics of hidden defects in 3D GPR results have not been well studied due to the complexity of urban road structures. Consequently, accurately detecting all subsurface defects in GPR survey areas remains challenging. In this paper, numerical models of road structures with typical types of hidden defects are constructed, and their reflection characteristics in GPR B-scan profiles and C-scan depth slices are analyzed. Then, the field results obtained using a car-mounted 3D GPR system are presented to confirm the reflection characteristics of these hidden defects. Finally, the reflections of manhole covers and underground pipelines, which are the most prevalent forms of clutter in GPR images in urban environments, are compared with those of hidden defects. The findings of this paper will facilitate the accurate interpretation of field 3D GPR data for road inspections and the mitigation of road collapse accidents.

GAN-Based Inversion of Crosshole GPR Data to Characterize Subsurface Structures

The crosshole ground-penetrating radar (GPR) technique is widely used to characterize subsurface structures, yet the interpretation of crosshole GPR data involves solving non-linear and ill-posed inverse problems. In this work, we developed a generative adversarial network (GAN)-based inversion framework to translate crosshole GPR images to their corresponding 2D defect reconstruction images automatically. This approach uses fully connected layers to extract global features from crosshole GPR images and employs a series of cascaded U-Net structures to produce high-resolution defect reconstruction results. The feasibility of the proposed framework was demonstrated on a synthetic crosshole GPR dataset created with the finite-difference time-domain (FDTD) method and real-world data from a field experiment. Our inversion network obtained recognition accuracy of 91.36%, structural similarity index measure (SSIM) of 0.93, and RAscore of 91.77 on the test dataset. Furthermore, comparisons with ray-based tomography and full-waveform inversion (FWI) suggest that the proposed method provides a good balance between inversion accuracy and efficiency and has the best generalization when inverting actual measured crosshole GPR data.

Combined CNN and RNN Neural Networks for GPR Detection of Railway Subgrade Diseases.

Vehicle-mounted ground-penetrating radar (GPR) has been used to non-destructively inspect and evaluate railway subgrade conditions. However, existing GPR data processing and interpretation methods mostly rely on time-consuming manual interpretation, and limited studies have applied machine learning methods. GPR data are complex, high-dimensional, and redundant, in particular with non-negligible noises, for which traditional machine learning methods are not effective when applied to GPR data processing and interpretation. To solve this problem, deep learning is more suitable to process large amounts of training data, as well as to perform better data interpretation. In this study, we proposed a novel deep learning method to process GPR data, the CRNN network, which combines convolutional neural networks (CNN) and recurrent neural networks (RNN). The CNN processes raw GPR waveform data from signal channels, and the RNN processes features from multiple channels. The results show that the CRNN network achieves a higher precision at 83.4%, with a recall of 77.3%. Compared to the traditional machine learning method, the CRNN is 5.2 times faster and has a smaller size of 2.6 MB (traditional machine learning method: 104.0 MB). Our research output has demonstrated that the developed deep learning method improves the efficiency and accuracy of railway subgrade condition evaluation.

A Modular Method for GPR Hyperbolic Feature Detection and Quantitative Parameter Inversion of Underground Pipelines

Ground penetrating radar (GPR) is widely used to inspect underground pipelines because it is non-destructive. When the scan line of GPR is perpendicular to the pipe, it will exhibit hyperbolic features in GPR B-scan images, which have no intuitive relationship with the geometric and physical parameters of the pipeline, making the interpretation of GPR images difficult. This paper proposes a modular detection and quantitative inversion method for the hyperbolic features in GPR B-scan images, which is divided into two steps. In the first step, the YOLOv7 object detection network is used to automatically detect the hyperbolic features in GPR images. In the second step, a two-stage curve fitting method is proposed based on the characteristics of the detection model. It uses a few key point annotations of the hyperbolic pattern and some parameters of the GPR system to quantitatively invert the depth and radius of pipes. Using the same hardware and data set, YOLOv7 achieves an 11.1% improvement in detection accuracy and an 18.2% improvement in speed compared to YOLOv5. The relative errors of the proposed method for the depth and radius of the synthetic data in homogeneous media are 0.6% and 4.4%, respectively, and 4.8% and 15% in non-homogeneous media. The relative error of the depth inversion of the measured data TU1208 is less than 10%. The results show that the method can effectively invert the depth and radius of underground pipelines and reduce the difficulty of GPR data interpretation.

Automatic analysis of railway ground penetrating radar: Using signal processing and machine learning approaches to assess railroad track substructure

Prior to track rehabilitation works on the French railway network, train-mounted ground penetrating radar (GPR) allows for fast and nondestructive data acquisition. Substructure condition evaluation is of paramount importance as it has an impact on the quality of the planned renewal works. The interpretation of GPR data remains a challenging and time-consuming task which requires expert intervention. This research seeks to enhance and automate the analysis of the GPR data using two different approaches. A signal processing approach based on entropy analysis determines the layer thicknesses, evaluates ballast fouling, and locates areas with water retention. Field comparisons showed concordances with the proposed method. The second approach relies on deep learning to detect mud pumping defects. It showed promising results for the detection of high intensity mud pumping.

A Multi-Path Encoder Network for GPR Data Inversion to Improve Defect Detection in Reinforced Concrete

Ground penetrating radar (GPR) has been extensively used in the routine inspection of reinforced concrete structures. However, the signatures in GPR images are reflected electromagnetic waves rather than their actual shapes. The interpretation of GPR data is a mandatory but time- and labor-consuming task. Furthermore, the rebars in the near-surface of concrete cause clutter in the GPR images, which hinders the interpretation of GPR data. This work presents a deep learning network to invert GPR B-scan images to permittivity maps of subsurface structures. The proposed network has a multi-path encoder which enables the network to leverage three kinds of GPR data: the original, migrated, and encoder–decoder-processed GPR data. Each type of processing method is designed to serve a different purpose: the original GPR images retain all the waveforms; the migration method intensifies the vertices of the subsurface anomalies; the encoder–decoder network suppresses rebar clutter and enhances the visibility of the defect echoes. The outputs of three processing methods are jointly used to interpret GPR B-scan images. We demonstrated the superiority of the proposed network by comparing it with a network with a single-path encoder. We also validated the proposed network with synthetic and experimental GPR data. The results indicate that the proposed network effectively reconstructs the defects in the reinforced concrete.

Analysis of 2D and 3D GPR data interpretation using continuous wavelet transforms: Case study from an archaeological test site

Ground-penetrating radar (GPR) is one of the most important techniques for obtaining high-resolution data in archaeological research, and it is becoming increasingly important. The continuous wavelet transform (CWT), which is non-numerical technique, gives an overcomplete representation of a signal by continuously varying the wavelet’s translation and scale parameters in the time series dataset. This paper focuses on the novel technique of integrating CWT and the wavelet transform maxima (WTM) to extract information from an archaeological test site in south-eastern China. For the characterization of archaeological features, we assessed the importance of dense and accurate data collection as well as GPR signal processing. The mathematical formulation and applicability of GPR attributes, particularly amplitude-based attributes, to identify and characterize archaeological buried targets are also discussed. GPR data is acquired using co-polarized and cross-polarized configurations with transverse-electric (TE) and transverse-magnetic (TM) broadside frequency plates at 100 and 200 MHz. Next, CWT was applied using six different wavelet levels, followed by amplitude comparison. The archaeological targets were successfully interpreted using peak amplitude and CWT. The proposed methodology has significantly improved data visualization and interpretation of GPR data, and it also gave us good results in identifying archaeological anomalies.

GPR Data Processing and Interpretation Based on Artificial Intelligence Approaches: Future Perspectives for Archaeological Prospection

Ground penetrating radar (GPR) is a well-established technique used in archaeological prospection and it requires a number of specialized routines for signal and image processing to enhance the data acquired and lead towards a better interpretation of them. Computer-aided techniques have advanced the interpretation of GPR data, dealing with a wide range of operations aiming towards locating, imaging, and diagnosis/interpretation. This article will discuss the novel and recent applications of machine learning (ML) and deep learning (DL) techniques, under the artificial intelligence umbrella, for processing GPR measurements within archaeological contexts, and their potential, limitations, and possible future prospects.

Phantom subsurface targets in ground-penetrating radar data

Ground-penetrating radar (GPR) has been a very effective tool for exploring the subsurface and the nondestructive testing of nonmetallic structures for the past 40–50 years. The traditional GPR data interpretation is built upon the innate bias that all signals emanate from within the ground and most GPR users are normally under the impression that energy mostly travels straight down leading to the perception that “targets” are beneath the measurement location. The response of features at the ground surface and above ground also is present in most data but not always consciously noted as contributing to the measurements. One class of responses from above-ground features is routinely called “airwaves” because they normally exhibit moveout velocities of air. Often, an above-ground source is not the first thing that comes to mind during data interpretation, unless the user is experienced. Even experienced users can occasionally be misled, as above-ground features are expected to reach the GPR receiver with the moveout velocity of air. Recent experience in some of our surveys has created concerns because the targets at or above the ground surface demonstrated ground wave moveout velocity, which eliminates one of the diagnostic tools. This paper explores this issue, identifies GPR signal paths, and suggests key factors to consider in field operations and data interpretation. To demonstrate the concepts described, we have used numerical modeling and field data sets.

3D Sedimentary Architecture of Sandy Braided River, Based on Outcrop, Unmanned Aerial Vehicle and Ground Penetrating Radar Data

Ground Penetrating Radar (GPR) is a geophysical method that uses antennas to transmit and receive high-frequency electromagnetic waves to detect the properties and distribution of materials in media. In this paper, geological observation, UAV detection and GPR technology are combined to study the recent sediments of the Yungang braided river study area in Datong. The application of the GPR technique to the description of fluvial facies and reservoir architecture and the development of geological models are discussed. The process of GPR detection technology and application includes three parts: GPR data acquisition, data processing and integrated interpretation of GPR data. The geological surface at different depths and scales can be identified by using different combinations of frequencies and antenna configurations during acquisition. Based on outcrop observation and lithofacies analysis, the Yandong Member of the Middle Jurassic Yungang Formation in the Datong Basin has been identified as a typical sandy braided river sedimentary system. The sandy braided river sandbody changes rapidly laterally, and the spatial distribution and internal structure of the reservoir are very complex, which has a very important impact on the migration and distribution of oil and gas as a reservoir. It is very important to make clear the characteristics of each architectural unit of the fluvial sand body and quantitatively characterize them. The architectural elements of the braided river sedimentary reservoir in the Datong-Yungang area can be divided into three types: Channel unit, bar unit and overbank assemblages. The geological radar response characteristics of different types of sedimentary units are summarized and their interfaces are identified. The channel sediments form a lens-shaped wave reflection with a flat at the top and convex-down at the bottom in the radar profile, and the angles of the radar reflection directional axes are different on both sides of the sedimentary interface. In the radar profile, the deposit of the unit bar is an upward convex reflection structure. The overbank siltation shows a weak amplitude parallel reflection structure. The flood plain sediments are distributed continuously and stably in the radar profile, showing weak reflection characteristics. Different sedimentary units are identified by GPR data and combined with Unmanned Aerial Vehicle (UAV) detection data, and the establishment of the field outcrop geological model is completed. The development pattern of the diara is clarified, and the swing and migration of the channel in different stages are identified.

Development of a workflow for processing ground-penetrating radar data from multiconcurrent receivers

Ground-penetrating radar (GPR) systems with multiconcurrent sampling receivers can rapidly acquire dense multioffset GPR data, which are not feasible using typical common-offset (CO) GPR systems with a single fixed offset transmitter-receiver pair. Multioffset GPR data from these new multiconcurrent receiver systems have the potential to be used to create detailed subsurface velocity models and enhanced reflection sections. These are important features that can improve qualitative and quantitative interpretation of GPR data. To realize these benefits and to deal with the large amount of multioffset data generated by these new systems, we have developed an automated and customized data processing workflow. There are three key algorithms that we have developed as part of our workflow, which is crucial for processing large volume, multioffset GPR data so as: first, to efficiently correct and manage time misalignments from multiconcurrent receivers; second, to carry out trace balancing of common-midpoint data for semblance analysis; and third, to automate the velocity analysis step. We showcase our processing workflow using two field data sets acquired using a multiconcurrent sampling receiver GPR system consisting of one transmitter and seven receivers. The field data were collected at two different locations: a site using a system with a 500 MHz center frequency and another site using a system with a 1000 MHz center frequency. We have determined, with both data sets, that our processing workflow could produce automated stacking velocity fields and enhanced zero-offset reflection cross sections. These benefits increase the information that can be used for interpretation (compared with conventional CO data) and can form the basis of further processing steps such as migration. As the cost of these multiconcurrent sampling receiver systems decreases over time, we anticipate their use, and the acquisition of dense multioffset GPR data, to become much more commonplace.

3D ground-penetrating radar attribute classification: A case study from a paleokarst breccia pipe in the Billefjorden area on Spitsbergen, Svalbard

Ground-penetrating radar (GPR) is a method that can provide detailed information about the near subsurface in sedimentary and carbonate environments. The classical interpretation of GPR data (e.g., based on manual feature selection) often is labor-intensive and limited by the experience of the interpreter. Novel attribute-based classification approaches, typically used for seismic interpretation, can provide faster, more repeatable, and less biased interpretations. We have recorded a 3D GPD data set collected across a paleokarst breccia pipe in the Billefjorden area on Spitsbergen, Svalbard. After performing advanced processing, we compare the results of a classical GPR interpretation to the results of an attribute-based classification. Our attribute classification incorporates a selection of dip and textural attributes as the input for a k-means clustering approach. Similar to the results of the classical interpretation, the resulting classes differentiate between undisturbed strata and breccias or fault zones. The classes also reveal details inside the breccia pipe that are not discerned in the classical interpretation. Using nearby outcropping breccia pipes, we infer that the intrapipe GPR facies result from subtle differences, such as breccia lithology, clast size, or pore-space filling.

The Zonation of Mountain Frozen Ground under Aspect Adjustment Revealed by Ground-Penetrating Radar Survey—A Case Study of a Small Catchment in the Upper Reaches of the Yellow River, Northeastern Qinghai–Tibet Plateau

Permafrost distribution is of great significance for the study of climate, ecology, hydrology, and infrastructure construction in high-cold mountain regions with complex topography. Therefore, updated high-resolution permafrost distribution mapping is necessary and highly demanded in related fields. This case study conducted in a small catchment in the northeast of the Qinghai Tibet Plateau proposes a new method of using ground-penetrating radar (GPR) to detect the stratigraphic structure, interpret the characteristics of frozen ground, and extract the boundaries of permafrost patches in mountain areas. Thus, an empirical–statistical model of mountain frozen ground zonation, along with aspect (ASP) adjustment, is established based on the results of the GPR data interpretation. The spatial mapping of the frozen ground based on this model is compared with a field survey dataset and two existing permafrost distribution maps, and their consistencies are all higher than 80. In addition, the new map provides more details on the distribution of frozen ground. In this case, the influence of ASP on the distribution of permafrost in mountain areas is revealed: the adjustment of ASP on the lower limit of continuous and discontinuous permafrost is 180–200 m, the difference in the annual mean ground temperature between sunny and shady slopes is up to 1.4–1.6 °C, and the altitude-related temperature variation and uneven distribution of solar radiation in different ASPs comprehensively affect the zonation of mountain frozen ground. This work supplements the traditional theory of mountain permafrost zonation, the results of which are of value to relevant scientific studies and instructive to engineering construction in this region.

Review of Machine Learning Algorithms for Automatic Detection of Underground Objects in GPR Images

Ground-penetrating radar (GPR) is a nondestructive tool that has gained popularity after giving promising results in different areas—such as utility engineering, transportation engineering, civil engineering, and geology—with relatively low cost. Even as the number of applications for GPR increases, the interpretation of GPR data is still challenging, in part due to varying ground conditions. Researchers are continuously working on the development of new analysis methods to address these challenges. Computer vision algorithms, including neural networks and convolution neural networks, have advanced significantly over the past decade, and researchers have utilized these algorithms to extract information from GPR images and thus improve the interpretation of GPR data. This paper presents a review of literature that employs computer vision and machine learning algorithms, such as YOLO V3, Viola–Jones, and AlexNet, for automatic extraction of information from GPR images. The uptake in the use of automatic detection algorithms for GPR is increased by the ability to rapidly quantify and locate buried targets that previously could only be identified by professionals with a high level of expertise and training.

3D Visualization Techniques for Analysis and Archaeological Interpretation of GPR Data

The non-invasive detection and digital documentation of buried archaeological heritage by means of geophysical prospection is increasingly gaining importance in modern field archaeology and archaeological heritage management. It frequently provides the detailed information required for heritage protection or targeted further archaeological research. High-resolution magnetometry and ground-penetrating radar (GPR) became invaluable tools for the efficient and comprehensive non-invasive exploration of complete archaeological sites and archaeological landscapes. The analysis and detailed archaeological interpretation of the resulting large 2D and 3D datasets, and related data from aerial archaeology or airborne remote sensing, etc., is a time-consuming and complex process, which requires the integration of all data at hand, respective three-dimensional imagination, and a broad understanding of the archaeological problem; therefore, informative 3D visualizations supporting the exploration of complex 3D datasets and supporting the interpretative process are in great demand. This paper presents a novel integrated 3D GPR interpretation approach, centered around the flexible 3D visualization of heterogeneous data, which supports conjoint visualization of scenes composed of GPR volumes, 2D prospection imagery, and 3D interpretative models. We found that the flexible visual combination of the original 3D GPR datasets and images derived from the data applying post-processing techniques inspired by medical image analysis and seismic data processing contribute to the perceptibility of archaeologically relevant features and their respective context within a stratified volume. Moreover, such visualizations support the interpreting archaeologists in their development of a deeper understanding of the complex datasets as a starting point for and throughout the implemented interactive interpretative process.