Frequency–Wavenumber Analysis of Deep Learning-based Super Resolution 3D GPR Images

Kang, Man-Sung; An, Yun‐Kyu

doi:10.3390/rs12183056

Cited by 22 publications

(5 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Arbitrary underground medium inhomogeneity and undesired measurement noises often disturb radar data interpretation [45]. A lake with a more homogeneous subsurface, such as Qinghai Lake, has a flatter ice surface, and this ensures better data collection when it enters the stable freeze-up period in winter.…”

Section: Discussionmentioning

confidence: 99%

Ice Thickness Assessment of Non-Freshwater Lakes in the Qinghai–Tibetan Plateau Based on Unmanned Aerial Vehicle-Borne Ice-Penetrating Radar: A Case Study of Qinghai Lake and Gahai Lake

Jin,

Yao,

Wei

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

Ice thickness has a significant effect on the physical and biogeochemical processes of a lake, and it is an integral focus of research in the field of ice engineering. The Qinghai–Tibetan Plateau, known as the Third Pole of the world, contains numerous lakes. Compared with some information, such as the area, water level, and ice phenology of its lakes, the ice thickness of these lakes remains poorly understood. In this study, we used an unmanned aerial vehicle (UAV) with a 400/900 MHz ice-penetrating radar to detect the ice thickness of Qinghai Lake and Gahai Lake. Two observation fields were established on the western side of Qinghai Lake and Gahai Lake in January 2019 and January 2021, respectively. Based on the in situ ice thickness and the propagation time of the radar, the accuracy of the ice thickness measurements of these two non-freshwater lakes was comprehensively assessed. The results indicate that pre-processed echo images from the UAV-borne ice-penetrating radar identified non-freshwater lake ice, and we were thus able to accurately calculate the propagation time of radar waves through the ice. The average dielectric constants of Qinghai Lake and Gahai Lake were 4.3 and 4.6, respectively. This means that the speed of the radar waves that propagated through the ice of the non-freshwater lake was lower than that of the radio waves that propagated through the freshwater lake. The antenna frequency of the radar also had an impact on the accuracy of ice thickness modeling. The RMSEs were 0.034 m using the 400 MHz radar and 0.010 m using the 900 MHz radar. The radar with a higher antenna frequency was shown to provide greater accuracy in ice thickness monitoring, but the control of the UAV’s altitude and speed should be addressed.

show abstract

Section: Discussionmentioning

confidence: 99%

Ice Thickness Assessment of Non-Freshwater Lakes in the Qinghai–Tibetan Plateau Based on Unmanned Aerial Vehicle-Borne Ice-Penetrating Radar: A Case Study of Qinghai Lake and Gahai Lake

Jin,

Yao,

Wei

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…A CNN-based SR algorithm has been applied to groundpenetrating radar (GPR). Specifcally, to reduce the analysis distortion of GPR, which is a nondestructive method, Kang and An [24] opted to decrease the amount of noise afecting the interpretation of data. Bae et al [25] proposed SrcNet for microcrack detection in large-scale bridges.…”

Section: Super-resolutionmentioning

confidence: 99%

Learning Structure for Concrete Crack Detection Using Robust Super-Resolution with Generative Adversarial Network

Kim

Shim

Kang

et al. 2023

Structural Control and Health Monitoring

View full text Add to dashboard Cite

To develop countermeasures against the inevitable aging of tunnels, accurate inspection is critical for ensuring stable tunnel services. Tunnels are large-scale infrastructures, whereas the cracks in such tunnels are small-scale objects, necessitating the development of methods capable of rapidly detecting cracks in a wide field of view. Therefore, this study proposes a new learning structure based on a segmentation network. This structure includes super-resolution and generative adversarial networks, which contribute to improved detection performance with robustness to input data. Furthermore, a method for the effective construction of training data for the application of the proposed structure is presented. Subsequently, the performance of the method over 1,606 crack images with randomly degraded qualities is evaluated. The proposed structure presents improved crack intersection over union and F1-scores of 63.686% and 77.811%, respectively, and low variances of 0.9008 and 0.5015 compared to the original structure. The results presented herein indicate the possible application of the proposed accurate condition inspection technology to tunnel maintenance in the future.

show abstract

“…Researchers have invested much effort in GPR-based subsurface inspection. A GPR can not provide 3D shape information, but a convoluted reflection image with cluttered signals, making it difficult to recognize subsurface defects automatically [14]. It is possible to detect subsurface objects using spatially informative GPR data, but interpreting GPR data automatically remains a challenge.…”

Section: Related Workmentioning

confidence: 99%

MV-GPRNet: Multi-View Subsurface Defect Detection Network for Airport Runway Inspection Based on GPR

et al. 2022

Remote Sensing

View full text Add to dashboard Cite

The detection and restoration of subsurface defects are essential for ensuring the structural reliability of airport runways. Subsurface inspections can be performed with the aid of a robot equipped with a Ground Penetrating Radar (GPR). However, interpreting GPR data is extremely difficult, as GPR data usually contains severe clutter interference. In addition, many different types of subsurface defects present similar features in B-scan images, making them difficult to distinguish. Consequently, this makes later maintenance work harder as different subsurface defects require different restoration measures. Thus, to automate the inspection process and improve defect identification accuracy, a novel deep learning algorithm, MV-GPRNet, is proposed. Instead of traditionally using GPR B-scan images only, MV-GPRNet utilizes multi-view GPR data to robustly detect regions with defects despite significant interference. It originally fuses the 3D feature map in C-scan data and the 2D feature map in Top-scan data for defect classification and localization. With our runway inspection robot, a large number of real runway data sets from three international airports have been used to extensively test our method. Experimental results indicate that the proposed MV-GPRNet outperforms state-of-the-art (SOTA) approaches. In particular, MV-GPRNet achieves F1 measurements for voids, cracks, subsidences, and pipes at 91%, 69%, 90%, and 100%, respectively.

show abstract

Frequency–Wavenumber Analysis of Deep Learning-based Super Resolution 3D GPR Images

Cited by 22 publications

References 43 publications

Ice Thickness Assessment of Non-Freshwater Lakes in the Qinghai–Tibetan Plateau Based on Unmanned Aerial Vehicle-Borne Ice-Penetrating Radar: A Case Study of Qinghai Lake and Gahai Lake

Ice Thickness Assessment of Non-Freshwater Lakes in the Qinghai–Tibetan Plateau Based on Unmanned Aerial Vehicle-Borne Ice-Penetrating Radar: A Case Study of Qinghai Lake and Gahai Lake

Learning Structure for Concrete Crack Detection Using Robust Super-Resolution with Generative Adversarial Network

MV-GPRNet: Multi-View Subsurface Defect Detection Network for Airport Runway Inspection Based on GPR

Contact Info

Product

Resources

About