Raw Ground Penetrating Radar Data Research Articles

Improving 3D Metric GPR Imaging Using Automated Data Collection and Learning-Based Processing

Ground Penetrating Radar (GPR) is one of the most important non-destructive evaluation (NDE) devices to detect subsurface objects (i.e., rebars, utility pipes) and reconstruct the underground scene. There are two challenges for GPR-based inspection, which are GPR data collection and 3D subsurface object imaging. To address these challenges, we first propose a robotic solution that automates the GPR data collection process with a free motion pattern. It facilitates the 3D metric GPR imaging by tagging the pose with GPR measurement in real-time. Moreover, to improve the 3D GPR imaging, we introduce a learning-based GPR data analysis method, which includes a noise removal module to clear the background noise in raw GPR data and a Convolutional Recurrent Neural Network (CRNN) to estimate the dielectric value of subsurface medium in each GPR B-scan data. We use both field and synthetic data to verify the proposed methods. Experimental results demonstrate that our proposed methods can achieve higher performance and faster processing speed in 3D GPR imaging than baseline methods.

Robotic Inspection of Underground Utilities for Construction Survey Using a Ground Penetrating Radar

Ground penetrating radar (GPR) is a very useful nondestructive evaluation (NDE) device for locating and mapping underground assets prior to digging and trenching efforts in construction. This paper presents a novel robotic system to automate the GPR data-collection process, localize underground utilities, and interpret and reconstruct the underground objects for better visualization, allowing regular nonprofessional users to understand the survey results. This system is composed of three modules: (1) an omnidirectional robotic data-collection platform that carries a RGB-D camera with inertial measurement unit (IMU) and a GPR antenna to perform automatic GPR data collection and tag each GPR measurement with visual positioning information at every sampling step, (2) a learning-based migration module to interpret the raw GPR B-scan image into a two-dimensional (2D) cross-section model of objects, and (3) a three-dimensional (3D) reconstruction module, i.e., 30.0% GPRNet, to generate underground utility model represented as fine 3D point cloud. Comparative studies were performed on synthetic data and field GPR raw data with various incompleteness and noise. Experimental results demonstrated that our proposed method achieves a higher GPR imaging accuracy in mean intersection over union (IoU) than the conventional back-projection (BP) migration approach, and 6.9%−7.2% less loss in Chamfer distance (CD) than point cloud model reconstruction baseline methods. The GPR-based robotic inspection provides an effective tool for civil engineers to detect and survey underground utilities before construction.

RNMF-Guided Deep Network for Signal Separation of GPR Without Labeled Data

The clutter encountered in the ground-penetrating radar (GPR) system severely obscures the visibility of subsurface objects, especially in the case of overlapping target responses and clutter. In this letter, a novel self-supervised learning strategy with dual-network architecture and pseudolabels is proposed. First, the dual-network consists of two subnetworks: one is to simulate the low-rank part, and another simulates the sparse part. Second, the raw GPR data are decomposed as the sum of low-rank and sparse matrices by robust nonnegative matrix factorization (RNMF), termed two pseudolabels. Then, these pseudolabels guide the two subnetworks to accurately reconstruct the target response and clutter trace by trace, respectively. Results based on simulated data by gprMax and real datasets demonstrate that the proposed method is effective in separating target response from clutter and can achieve a similar effect as RNMF in less time without any prior information.

A pipeline for enhanced multimodal 2D imaging of concrete structures

We present an imaging pipeline to achieve enhanced images of the interior of concrete from ground penetrating radar (GPR) and ultrasonic echo array (UEA) measurements. This work lays the foundation for an advanced yet practical imaging tool to assess concrete structures. Specifically, we propose an enhanced two-dimensional (2D) total focusing method (XTFM) to reconstruct images from raw GPR and UEA data. The proposed XTFM algorithm integrates total focusing method (TFM) and synthetic aperture focusing technique (SAFT) concepts to post-process large independent and interelement measurements from both modalities in a computationally efficient way. Furthermore, we introduce a novel 2D image fusion algorithm using wavelet multilevel decomposition and an NDT knowledge-based rule to fuse GPR and UEA images. We then compare our algorithm with conventional fusion algorithms such as averaging, maximum, and product. The results from three laboratory concrete reference specimens are evaluated in detail. The fused images are compared with each other as well as benchmarked with the original GPR and UEA images. The output image obtained from our proposed pipeline is an enhanced 2D image of the interior of concrete structures that eases interpretation by a human inspector as well as it has the potential to improve interpretation by computer vision and image analysis algorithms.

GPR clutter reduction by multi-resolution based tensor RPCA

ABSTRACT A new clutter reduction method which utilizes the multi-resolution and multi-directional information of the ground-penetrating radar (GPR) image is proposed. Sub-images obtained by stationary wavelet transform (SWT) or nonsubsampled counterlet transform (NSCT) are cast into a tensor structure presenting higher information compared to the spatial input data. A tensor-robust principal component analysis (TRPCA) algorithm is used for low-rank and sparse decomposition (LRSD) followed by inverse transform of the sparse tensor component to provide the clutter reduction results. The proposed methods TRPCA-SWT and TRPCA-NSCT are compared both visually and quantitatively to robust principal component analysis (RPCA) and TRPCA-bandpass filter (TRPCA-BPF), which employ the spatial raw GPR data and outputs of simple low-pass and high-pass filters respectively. Visual and quantitative results demonstrate that the clutter reduction performance increases when a higher number of scales and directions are used prior to the LRSD decomposition. Moreover, one of the proposed methods, TRPCA-NSCT, removes the background noise more efficiently due to its higher multi-resolution and multi-direction investigation capability, increasing the performance of the target detection algorithms.

Application of Ground Penetrating Radar in Detecting Deeply Embedded Reinforcing Bars in Pile Foundation

Ground penetrating radar (GPR) has been widely used for nondestructive testings in civil engineering. However, the GPR has not been adequately applied in detecting deeply embedded reinforcing bars, which is usually difficult to be revealed in radar image due to the wave interference and attenuation in large depth penetration. This study presents a new approach for the GPR detection of deeply embedded reinforcing bars in the reinforced concrete pile foundation. The aim of the GPR survey is to determine the existence and the depth of internal reinforcing bars in the pile foundation for solving engineering dispute. Low centre frequency antenna was used in GPR field testing to obtain the reflected raw data. Optimized procedures of digital filtering techniques were applied to process the GPR raw data. The deeply embedded reinforcing bars are revealed in the radar image after the field testing and postprocessing procedures. The depth of the reinforcing bars was estimated based on the hyperbola match method. The GPR test results were validated by the excavation of the pile foundation. The low centre frequency antenna has been found to be essential to obtain the reflected wave signals of deeply embedded reinforcing bars. The optimized processing procedures is useful to identify and display the reinforcing bars in radar image. The combination of low centre frequency antenna and the postprocessing procedures make the detection of deeply embedded reinforcing bars feasible. The proposed GPR testing method has been found to be effective to estimate the depth of deeply embedded reinforcing bars, which provides the key information for solving engineering dispute.

3-D Multistatic Ground Penetrating Radar Imaging for Augmented Reality Visualization

Ground penetrating radar (GPR) is a useful instrument for smarter infrastructure applications, in particular, for the localization and mapping of underground infrastructure and other subsurface assets, due to its ability to sense metallic and nonmetallic buried objects. For instance, air-coupled, multistatic GPR could potentially be employed to quickly produce subsurface maps for public and private stakeholders, enabling rational and more efficient planning of underground infrastructure inspection, maintenance, and construction. An application of interest in such context is a faster identification of underground utilities location and depth by innovative data visualization methods, such as augmented reality. A 3-D model of the subsurface asset is desirable for such applications. However, raw GPR data is often hard to interpret. Imaging algorithms are applied to improve GPR data readability and signal-to-noise ratio by focusing the spread energy. Here, a processing pipeline that takes raw 3-D multistatic GPR data as input and yields a 3-D model as output is proposed. Initially, a 3-D back-projection algorithm is applied to air-coupled, multistatic GPR data to recover buried target localization. An enhancement filter, tailored for tubular structures, is applied to reduce background noise and highlight structures of interest in the 3-D image. This process is successfully applied to three laboratory scenarios of plastic buried targets with different sizes and shapes.

Metal shape classification of buried object using multilayer perceptron neural network in GPR data

Ground penetrating radar (GPR) is a non-destructive tool for subsurface investigation becomes an emerging trend in many applications especially in detecting infrastructure utilities. However, interpreting of GPR data in terms of target’s geometry is remained a challenging task due to the existence intra/inter-class variations and subtle changes of hyperbolic signature pattern in GPR data. To addressed this issue, this paper proposes to classify the basic shape (i.e. cubic, cylindrical and disc) of buried object using multilayer perceptron neural network. The GPR raw data firstly is pre-processed based on A-scan and B- scan GPR data. Then, the combined statistical features with gray level co-occurrence matrix (GLCM) and Hu moment invariants are extracted from A-scan and B-scan. The extracted features are then fed as input to multilayer perceptron to classify the shapes (cubic, cylindrical and disc) of buried object. Based on the experiment conducted, the highest recognition rate has been achieved is about 83.3%. Thus, the proposed method of using combined statistical with GLCM and Hu moment invariants features to classify shape of buried object using multilayer perceptron is promising.

Improved Clutter Removal in GPR by Robust Nonnegative Matrix Factorization

The clutter encountered in the ground-penetrating radar (GPR) system severely decreases the visibility of subsurface objects, thus highly degrading the performance of the target detection algorithms. This letter presents a new clutter removal method based on nonnegative matrix factorization (NMF). The raw GPR data are represented as the sum of low-rank and sparse matrices, which correspond to the clutter and target components, respectively. The low-rank and sparse decomposition is performed using a robust version of NMF called RNMF. Although similar to the robust principal component analysis (PCA) (RPCA), which is recently widely used in image processing applications as well as in GPR, the proposed method is faster and has enhanced results. The state-of-the-art clutter removal methods, morphological component analysis (MCA), RPCA, besides the conventional PCA, have been included for comparison for both simulated and real data sets. The visual and quantitative results demonstrate that the proposed RNMF method outperforms the others. Moreover, it is 25 times faster than the RPCA for the given regularization parameter values.

An Efficient And Adaptive Compression Algorithm Based On DCT Transform For GPR Raw Data

Ground Penetrating Radar(GPR) creates large amounts of data which causes serious problems for storage and transmission. In this paper, a novel compression algorithm for GPR raw data based on DCT transform and inter-frame difference is proposed. The algorithm first conducts DCT transform on the GPR raw data, and differentiates the adjacent frame data in DCT domain to reduce the variance and dynamic range. Then the differential data is evenly divided into several blocks, and different bit Lloyd-Max quantizers are selected to quantify different sub-block data through an adaptive bit allocation mechanism, thereby obtaining a higher compression gain. The compression experiments with satellite-borne GPR, aircraft-borne GPR and vehicle-borne GPR raw data confirmed the effectiveness of the compression algorithm.

Improving estimates of buried pipe diameter and infilling material from ground-penetrating radar profiles with full-waveform inversion

Ground-penetrating radar (GPR) is a widely used tool for the detection and location of buried utilities. Buried pipes generate characteristic diffraction hyperbolas in raw GPR data. Current methods for analyzing the shapes and timing of the diffraction hyperbolas are very effective for locating pipes, but they are less effective for determining the diameter of the pipes, particularly when the pipes are smaller than the radar wavelengths, typically a few tens of centimeters. A full-waveform inversion (FWI) method is described for improving estimates of the diameter of a pipe and confirming the infilling material (air/water/etc.) for the simple case of an isolated diffraction hyperbola on a profile run perpendicular to a pipe with antennas in broadside mode (parallel to the pipe). The technique described here can improve a good initial guess of the pipe diameter (within 30%–50% of the true value) to a better estimate (less than approximately 8% misfit). This method is developed by combining two freely available software packages with a deconvolution method for GPR effective source wavelet estimation. The FWI process is run with the PEST algorithm (model-independent parameter estimation and uncertainty analysis). PEST iteratively calls the gprMax software package for forward modeling of the GPR signal as the model for the pipe and surrounding soil is refined.

Protective effect and treatment of cucurbitacin B on mice with sepsisassociated acute kidney injury

Ground Penetrating Radar (GPR) is vastly utilized for non-invasive study of man-made structures, particularly for determining depth of pipes which are buried under the ground. However, shallow objects might obscure GPR raw data which is reflected from deeper ones. The current work suggests a signal processing method known as the parallel lifting-based Canonical Signed Digit (CSD) 2D Discrete Wavelet Transform (DWT) infrastructure is suggested to realize a multiplier less architecture with lesser hardware complexity, smaller area and power consumption. The flipping structure of the lifting scheme can be recognized to reduce the longest critical path. CSD based structure is demonstrated with only adders and free of multipliers. A stripe based scanning method is adopted in order to achieve an efficient memory. JPEG 2000 lossy 9/7 filter is structured and the same scanning method with CSD is used to design 6/10 filter to contribute the evidence for the proposed methodology. In the existing method, memory efficiency is achieved with less speed. The new architecture is proposed using CSD multiplier with more speed. The filter coefficients are multiplied by 256 and are converted to integer form and then CSD representation is considered. The suggested architecture provides multiplier-less infrastructure for DWT utilizing CSD. This infrastructure is appropriate for high speed online applications with less area and power consumption. For an N× (N+1) image, the suggested CSD based lifting infrastructure utilizes solely 3N temporal memory, 2S transposition memory as well as Tm+Ta critical path.

Estimating active layer thickness and volumetric water content from ground penetrating radar measurements in Barrow, Alaska.

Ground penetrating radar (GPR) has emerged as an effective tool for estimating active layer thickness (ALT) and volumetric water content (VWC) within the active layer. In August 2013, we conducted a series of GPR and probing surveys using a 500 MHz antenna and metallic probe around Barrow, Alaska. We collected about 15 km of GPR data and 1.5 km of probing data. Here, we describe the GPR data processing workflow from raw GPR data to the estimated ALT and VWC. We include the corresponding uncertainties for each measured and estimated parameter. The estimated average GPR‐derived ALT was 41 cm, with a standard deviation of 9 cm. The average probed ALT was 40 cm, with a standard deviation of 12 cm. The average GPR‐derived VWC was 0.65, with a standard deviation of 0.14.

Temporal Monitoring of the Soil Freeze-Thaw Cycles over a Snow-Covered Surface by Using Air-Launched Ground-Penetrating Radar

We tested an off-ground ground-penetrating radar (GPR) system at a fixed location over a bare agricultural field to monitor the soil freeze-thaw cycles over a snow-covered surface. The GPR system consisted of a monostatic horn antenna combined with a vector network analyzer, providing an ultra-wideband stepped-frequency continuous-wave radar. An antenna calibration experiment was performed to filter antenna and back scattered effects from the raw GPR data. Near the GPR setup, sensors were installed in the soil to monitor the dynamics of soil temperature and dielectric permittivity at different depths. The soil permittivity was retrieved via inversion of time domain GPR data focused on the surface reflection. Significant effects of soil dynamics were observed in the time-lapse GPR, temperature and dielectric permittivity measurements. In particular, five freeze and thaw events were clearly detectable, indicating that the GPR signals respond to the contrast between the dielectric permittivity of frozen and thawed soil. The GPR-derived permittivity was in good agreement with sensor observations. Overall, the off-ground nature of the GPR system permits non-invasive time-lapse observation of the soil freeze-thaw dynamics without disturbing the structure of the snow cover. The proposed method shows promise for the real-time mapping and monitoring of the shallow frozen layer at the field scale.

Estimating Features of Underground Utilities: Hybrid GPR/GPS Approach

Ground penetrating radar (GPR) is capable of detecting, locating, and characterizing underground utilities in a non-destructive manner. However, processing raw GPR scans to estimate the buried depth and radius of underground utilities remains a challenge in practice, which can be attributed to two main constraints: (1) the necessity of the GPR scanning trajectory being perpendicular to the centerline of the utility to be scanned and (2) the requirement of knowing either the velocity of the GPR electromagnetic (EM) waves traveling through heterogeneous soils or the range of the utility’s radius. This paper presents a novel method to overcome the two limitations by simultaneously estimating the radius and the buried depth of underground utilities based on GPR scans and auxiliary global positioning system (GPS) data. First, GPR scans are pre-processed to extract raw data points. Thereafter, a generic hyperbola equation is proposed to model GPR raw data, which incorporates the relative angles between buried utilities and GPR scanning trajectories. The very important point (VIP) algorithm then is developed to estimate the radius and depth utilizing auxiliary GPS data. The proposed new method was validated using GPR field scans obtained under various settings and was found to increase the accuracy of estimating the buried depth and radius under a general scanning condition (i.e., when both the EM wave velocity and the range of the radius are unknown and the GPR scanning trajectory is not perpendicular to the centerline of the buried utility).

Performance evaluation of the neural networks for moisture detection using GPR

Ground penetrating radar (GPR) is a highly researched area; however, despite this, there is a lack of knowledge about the well-known problem of moisture distorting the results of GPR surveys. This research analyses the results of a GPR survey on a Case Study Bridge structure in order to analyse this effect, specifically when checking for the positioning of rebar. The expected distortions of the GPR results due to the presence of moisture were indeed present, as further evidenced by subsequent destructive testing and velocity analysis. Furthermore, neural networks were also utilised to detect moisture ingress from the GPR raw data.

GPR Raw-Data Order Statistic Filtering and Split-Spectrum Processing to Detect Moisture

Considerable research into the area of bridge health monitoring has been undertaken; however, information is still lacking on the effects of certain defects, such as moisture ingress, on the results of ground penetrating radar (GPR) surveying. In this paper, this issue will be addressed by examining the results of a GPR bridge survey, specifically the effect of moisture in the predicted position of the rebars. It was found that moisture ingress alters the radargram to indicate distortion or skewing of the steel reinforcements, when in fact destructive testing was able to confirm that no such distortion or skewing had occurred. Additionally, split-spectrum processing with order statistic filters was utilized to detect moisture ingress from the GPR raw data.

Noise Attenuation from GPR Data Using Wavelet Transform and Artificial Neural Network

In this paper, a tentative of noise attenuation from the 3D Ground Penetrating Radar (GPR) data using the wavelet transform and the Multilayer Perceptron neural network model is implanted. GPR data recorded in Algerian Sahara are filtered firstly using the discrete wavelet transform, after that a MLP machine with three layers is trained in a supervised learning mode, the input is an extracted profile from the raw 3D GPR data and the output consists of the GPR data of the same profile but after filtering. The estimated weights of connection are used to propagate the remaining non-filtered data through the implanted machine, the calculated output consists to the filtered GPR data. Comparison between the filtered data using the MLP machine and the continuous wavelet transform shows that the neural network machine can be used for S/N ratio improvement of the noisy GPR data.

Structure and genesis of the Thabor rock glacier (Northern French Alps) determined from morphological and ground-penetrating radar surveys

Landform analysis and ground-penetrating radar (GPR) were used to investigate the Thabor rock glacier, in the Northern French Alps. The surface features of the rock glacier were classified and described, with emphasis on massive ice exposures. The retreat of the former Thabor glacier since the Little Ice Age (LIA) was documented through an analysis of historical sources, and recent movements of the rock glacier were inferred from orthophoto-based measurements. Two-dimensional (2-D) models of the radar wave velocity were derived from the raw GPR data, using the numerous diffraction hyperbolae for local determinations of the velocity and kriging interpolation techniques. Subsequently, the profiles were migrated through a 2-D Kirchhoff migration method using the interpolated velocities. The 2-D velocity models exhibit pronounced spatial variations and, in several locations, high values (> 0.15 m ns −1) potentially corresponding to massive ice. On the other hand, while the migrated profiles show numerous layers, the internal stratigraphy of the rock glacier is dominated by a few prominent internal boundaries. The integration of morphology, radar wave velocity, and internal stratigraphy allowed us to identify the main structural units of the rock glacier as well as to explain its genesis: the rock glacier was formed by the imbrication of a massive ice core, originating from the retreat of the former Thabor glacier since the LIA, into pre-existing glacial deposits.

Buried pipe detection by ground penetrating radar using the discrete wavelet transform

Ground penetrating radar (GPR) is widely used for non-invasive examination of man-made structures, especially to determine the depth of pipes buried underground. Unfortunately, shallower objects may obscure GPR raw data that is reflected from deeper ones. This study introduces a signal processing technique, called the discrete wavelet transform (DWT), to filter and enhance the GPR raw data in order to obtain higher quality profile images. Laboratory experiments were conducted and the locations of buried pipes under different conditions were analyzed. The buried pipes were made of plastic and metal, and both single and two parallel horizontal pipes are discussed. The experimental results indicate that the DWT profiles can provide more information than the traditional GPR profile. The images of the diameter and position of pipes, even two pipes of different materials and in horizontal alignment, can be enhanced by using the DWT profile.