• 한국지구물리탐사학회:학술대회논문집
• /
• 2006.06a
• /
• pp.265-270
• /
• 2006

High-frequency electromagnetic (EM) wave propagation phenomena associated with borehole ground-penetrating radar (GPR) surveys are complex. To improve the understanding of governing physical processes, we present a finite-difference time-domain solution of Maxwell's equations in cylindrical coordinates. This approach allows us to model the full EM wavefield associated with borehole GPR surveys. The algorithm can be easily implemented perfectly matched layers for absorbing boundaries, frequency-dependent media, and finite-length transmitter antenna.

• Geophysics and Geophysical Exploration
• /
• v.26 no.4
• /
• pp.211-228
• /
• 2023

Ground-penetrating radar (GPR) surveys are commonly used to monitor embankments, which is a nondestructive geophysical method. The results of GPR surveys can be complex, depending on the situation, and data processing and interpretation are subject to expert experiences, potentially resulting in false detection. Additionally, this process is time-intensive. Consequently, various studies have been undertaken to detect cavities in GPR survey data using deep learning methods. Deep-learning-based approaches require abundant data for training, but GPR field survey data are often scarce due to cost and other factors constaining field studies. Therefore, in this study, a deep- learning-based model was developed for embankment GPR survey cavity detection using data augmentation strategies. A dataset was constructed by collecting survey data over several years from the same embankment. A you look only once (YOLO) model, commonly used in computer vision for object detection, was employed for this purpose. By comparing and analyzing various strategies, the optimal data augmentation approach was determined. After initial model development, a stepwise process was employed, including box clustering, transfer learning, self-ensemble, and model ensemble techniques, to enhance the final model performance. The model performance was evaluated, with the results demonstrating its effectiveness in detecting cavities in embankment GPR survey data.

• Geophysics and Geophysical Exploration
• /
• v.10 no.1
• /
• pp.69-76
• /
• 2007

High-frequency electromagnetic (EM) wave propagation associated with borehole ground-penetrating radar (GPR) is a complicated phenomenon. To improve the understanding of the governing physical processes, we employ a finite-difference time-domain solution of Maxwell's equations in cylindrical coordinates. This approach allows us to model the full EM wavefield associated with crosshole GPR surveys. Furthermore, the use of cylindrical coordinates is computationally efficient, correctly emulates the three-dimensional geometrical spreading characteristics of the wavefield, and is an effective way to discretise explicitly small-diameter boreholes. Numerical experiments show that the existence of a water-filled borehole can give rise to a strong waveguide effect which affects the transmitted waveform, and that excitation of this waveguide effect depends on the diameter of the borehole and the length of the antenna.

• Tunnel and Underground Space
• /
• v.5 no.2
• /
• pp.114-122
• /
• 1995

The principle and applications of GPR(Ground Penetrating Radear) are familiar to engineering geologists and geophsicists as very attractive technique for continuous high resolution images of the subsurface. However, the main limitation of GPR is obviously related to presence of clayey or silty conductive soils, resulting in complete attenuation of radar signals. This difficulty gives hesitation for the exploration of the deeper targets for example detecting bedrock, particularly in Korean situation that most regions have conductive overburden. In order to prove usefulness of geological survey with GPR in that situation, the technique was tried to investigate depth of bedrock under thick conductive overburden and the other geolocgical informations for the constructionof foundation in the Dongbu apartment site, Kimhae. The reflection patterns on the processed GPR sections are well correlated with the geotechnical units-bedrock, alluvium, landfill unit and their internal layer-boundaries of boring data before GPR survey, except upper contact of bedrock. The isopach maps of the geotechnical units for the 3-D interpretations are made from GPR sections. The maps provided useful geological information that bedrock was distributed as plain and valley with 22~27m depth under alluvium unit (this depth is 5~8 m deeper than drill log) and sedimentary layers subsided and bended along growth fault with NNE strike/15$^{\circ}$SE dip in alluvium unit.

• Geophysics and Geophysical Exploration
• /
• v.7 no.1
• /
• pp.93-98
• /
• 2004

The sand-covered Muweilah archaeological site in the United Arab Emirates (UAE) is a unique Iron Age site, and has been subject to intensive investigations. However, excavations are time consuming and may require twenty years to complete. Thus geophysical surveys were undertaken with the objective of characterising the site more expeditiously. This paper presents preliminary results of these surveys. Ground penetrating radar (GPR) was tested as a primary imaging tool, with an ancillary shallow time domain EM (MetalMapper) system. Dense 3D GPR datasets were migrated to produce horizontal (plan view) depth slices at 10 cm intervals, which is conceptually similar to the archaeologists' excavation methodology. The objective was to map all features associated with anthropogenic activity. This required delineating extensive linear and planar features, which could represent infrastructure. The correlation between these and isolated point reflectors, which could indicate anthropogenic activity, was then assessed. Finally, MetalMapper images were used to discriminate between metallic and non-metallic scatterers. The moderately resistive sand cover allowed GPR depth penetration of up to 5 m with a 500 MHz system. GPR successfully mapped floor levels, walls, and isolated anthropogenic activity, but crumbling walls were difficult to track in some cases. From this study, two possible courtyard areas were recognised. The MetalMapper was less successful because of its limited depth penetration of 50 cm. Despite this, the system was still useful in detecting modem-day ferruginous waste and bronze artefacts. The results (subject to ongoing ground-truthing) indicated that GPR was optimal for sites like Muweilah, which are buried under a few metres of sand. The 3D survey methodology proved essential to achieve line-to-line correlation for tracking walls. In performing the surveys, a significant improvement in data quality ensued when survey areas were flattened and de-vegetated. Although MetalMapper surveys were not as useful, they certainly indicated the value of including other geophysical data to constrain interpretation of complex GPR features.

• 한국지구물리탐사학회:학술대회논문집
• /
• 2004.08a
• /
• pp.8-21
• /
• 2004

Examples of GPR survey results at a variety of archaeological sites are presented. Several new analyses which include static corrections for the tilt of the GPR antenna are shown for imaging of burial mounds with significant topography. Example archaeological site plans developed from GPR remote sensing of Roman and Japanese sites are given. The first completely automated GPR survey, using only Global Positioning Satellite navigation to create 3D data volumes, is employed for a site in Louisiana to detect lost graves of the Choctaw Indian Tribe.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.20 no.2
• /
• pp.86-93
• /
• 2016

To find the buried pipes and cavities, GPR detection were proceed by the type and depth of underground pipes and cavities buried in the Chamber. In the case of asphalt pavement and non-pavement, the exploration of buried pipe were easy than the concrete and reinforced concrete pavement. In the case of air cavity, the buried depth of 1 m was evaluated as the detection was possible.

• Geophysics and Geophysical Exploration
• /
• v.13 no.1
• /
• pp.51-60
• /
• 2010

To determine subglacial topography and internal features of the Fourcade Glacier on King George Island in Antarctica, helicopter-borne and ground-towed ground-penetrating radar (GPR) data were recorded along four profiles in November 2006. Signature deconvolution, f-k migration velocity analysis, and finite-difference depth migration applied to the mixed-phase, single-channel, ground-towed data, were effective in increasing vertical resolution, obtaining the velocity function, and yielding clear depth images, respectively. For the helicopter-borne GPR, migration velocities were obtained as root-mean-squared velocities in a two-layer model of air and ice. The radar sections show rugged subglacial topography, englacial sliding surfaces, and localised scattering noise. The maximum depth to the basement is over 79m in the subglacial valley adjacent to the south-eastern slope of the divide ridge between Fourcade and Moczydlowski Glaciers. In the ground-towed profile, we interpret a complicated conduit above possible basal water and other isolated cavities, which are a few metres wide. Near the terminus, the GPR profiles image sliding surfaces, fractures, and faults that will contribute to the tidewater calving mechanism forming icebergs in Potter Cove.

• Geophysics and Geophysical Exploration
• /
• v.25 no.4
• /
• pp.189-200
• /
• 2022

Ground subsidence on urban roads is a social issue that can lead to human and property damages. Therefore, it is crucial to detect underground cavities in advance and repair them. Underground cavity detection is mainly performed using ground penetrating radar (GPR) surveys. This process is time-consuming, as a massive amount of GPR data needs to be interpreted, and the results vary depending on the skills and subjectivity of experts. To address these problems, researchers have studied automation and quantification techniques for GPR data interpretation, and recent studies have focused on deep learning-based interpretation techniques. In this study, we described a hyperbolic event detection process based on deep learning for GPR data interpretation. To demonstrate this process, we implemented a series of algorithms introduced in the preexisting research step by step. First, a deep learning-based YOLOv3 object detection model was applied to automatically detect hyperbolic signals. Subsequently, only hyperbolic signals were extracted using the column-connection clustering (C3) algorithm. Finally, the horizontal locations of the underground cavities were determined using regression analysis. The hyperbolic event detection using the YOLOv3 object detection technique achieved 84% precision and a recall score of 92% based on AP50. The predicted horizontal locations of the four underground cavities were approximately 0.12 ~ 0.36 m away from their actual locations. Thus, we confirmed that the existing deep learning-based interpretation technique is reliable with regard to detecting the hyperbolic patterns indicating underground cavities.

• 한국지구물리탐사학회:학술대회논문집
• /
• 2001.09a
• /
• pp.94-103
• /
• 2001

An experiment was conducted using a sand and gravel-filled tank model, to investigate the influence on the GPR response of vadose zone gasoline vapor phase effects and residual gasoline distributed by a fluctuating water table. After background GPR measurements were made with only water in the tank, gasoline was injected into the bottom of the model tank to simulate a subsurface discharge from a leaking pipe or tank. Results from the experiment show the sensitivity of GPR to the changes in the moisture content and its effectiveness for monitoring minor fluctuation of the water table. The results also demonstrate a potential of GPR for detecting possible vapor phase effects of volatile hydrocarbons in the vadose zone as a function of time, and for detecting the effects of residual phase of hydrocarbons in the water saturated system. In addition, the results provide the basis for a strategy that has the potential to successfully detect and delineate LNAPL contamination at field sites where zones of residual LNAPL in the water saturated system are present in the subsurface.