Automatic Recognition of Underground Pipelines using Ground Penetrating Radar Images
DOI:
https://doi.org/10.69980/ajpr.v28i5.821Keywords:
YOLOV8, Ground Penetrating Radar (GPR), Roboflow Universe, identifying subsurface, annotated.Abstract
The detection of underground locations and hidden pipelines is essential in subsurface engineering, utility mapping, and geophysical exploration. Conventional techniques dependent on manual analysis of Ground Penetrating Radar (GPR) data are labor-intensive and susceptible to human error. The current study introduces an automated methodology applying the YOLOv8 deep learning model for real-time item detection in GPR photos. A highly selected and annotated dataset obtained from the Roboflow Universe platform, comprising 1,474 GPR images annotated with bounding boxes for underground gaps and pipelines, was utilized for model training and assessment. The dataset included various subsurface circumstances, such as differences in soil type, moisture content, and buried depth, which enhanced the model's generality. Preprocessing involved image standardization and augmentation methods, including flipping, rotation, and brightness modifications to enhance adaptability. The trained model exhibited outstanding accuracy, attaining a precision of 0.90, a recall of 0.95, and a mean Average Precision (mAP50) of almost 0.95. The results show the efficacy of the YOLOv8 model in precisely detecting subsurface anomalies, presenting significant opportunities for enhancing the efficiency and reliability of GPR data analysis. This study focuses on the efficacy of open-source datasets and effective object detection algorithms in automating necessary tasks in underground infrastructure assessment.
References
1. Panjami, K., and Anbazhagan, P. (2025). Underground urban cavities: Consequences and identification through an integrated approach. In Underground Urban Cavities: Consequences and Identification Through an Integrated Approach (pp. 77–89). Springer.
2. Huang, S., Zhu, G., Tang, J., Li, W., and Fan, Z. (2025). Multi-perspective semantic segmentation of ground penetrating radar images for pavement subsurface objects. IEEE Transactions on Intelligent Transportation Systems.
3. Küçükdemirci, M., and Sarris, A. (2022). GPR data processing and interpretation based on artificial intelligence approaches: Future perspectives for archaeological prospection. Remote Sensing, 14(14), 3377.
4. Branco, D. R. A. (2024). Subsurface characterization using GPR data by means of image segmentation with YOLO-v8.
5. Han, Y., Shen, Y., Xu, J., Kariuki, J. M., and Li, Z. (2025). Research on automatic recognition technology of underground voids and pipelines in ground penetrating radar images. Sensing and Imaging, 26(1), 1–17.
6. Safaan, M. H., Metawie, M., and Marzouk, M. (2025). Subsurface utility detection and augmented reality visualization using GPR and deep learning. Automation in Construction, 176, 106299.
7. Zhang, C., Mousavi, A. A., Masri, S. F., and Gholipour, G. (2024). The state-of-the-art on time-frequency signal processing techniques for high-resolution representation of nonlinear systems in engineering. Archives of Computational Methods in Engineering, 1–22.
8. Li, J., Zhang, Z., and He, H. (2018). Hierarchical convolutional neural networks for EEG-based emotion recognition. Cognitive Computation, 10, 368–380.
9. Li, H., Gao, R., Sun, F., Wang, Y., and Ma, B. (2025). A prestressed concrete cylinder pipe broken wire detection algorithm based on improved YOLOv5. Sensors, 25(3), 977.
10. Hussain, M. (2024). Yolov5, yolov8 and yolov10: The go-to detectors for real-time vision. arXiv preprint arXiv:2407.02988.
11. Yang, Y., Li, Q., Li, H., Ji, H., Zeng, L., Wu, G., Wang, H., and Huang, Y. (2024). Accurate and automated detection of fractures in borehole image using YOLO-V8 model. Petroleum Science and Technology, 1–19.
12. Shahria, M. T., and Rahman, M. H. (2024). Activities of daily living object dataset: Advancing assistive robotic manipulation with a tailored dataset. Sensors, 24(23), 7566.
13. Guo, Q., Yang, P., Wu, R., and Zhang, Y. (2024). Numerical modeling of GPR for underground multi-pipes detection by combining GprMax and deep learning model. Progress in Electromagnetics Research M, 128.
14. Yao, Q., Zhuang, D., Feng, Y., Wang, Y., and Liu, J. (2024). Accurate detection of brain tumor lesions from medical images based on improved YOLOv8 algorithm. IEEE Access.
15. Wang, N., Zhang, Z., Hu, H., Li, B., and Lei, J. (2024). Underground defects detection based on GPR by fusing simple linear iterative clustering phash (SLIC-Phash) and convolutional block attention module (CBAM)-YOLOv8. IEEE Access, 12, 25888–25905.
16. Dou, Q., Wei, L., Magee, D. R., and Cohn, A. G. (2016). Real-time hyperbola recognition and fitting in GPR data. IEEE Transactions on Geoscience and Remote Sensing, 55(1), 51–62.
17. Liu, Y., Yuan, D., Song, C., Xu, T., and Fan, D. (2024). Hyperbolic attention-driven deep networks for enhanced GPR imaging of underground pipelines. IEEE Transactions on Geoscience and Remote Sensing.
18. Mojahid, A., Ouai, D. E., Amraoui, K. E., EL-Hami, K., Aitbenamer, H., Verrelst, J., and Barone, P. M. (2025). Lightweight CNN model for automatic detection and depth estimation of subsurface voids using GPR B-scan data. Natural Hazards Research.
19. Zhang, Y.-D., Wang, Q., Wu, C.-J., Wang, X.-N., Zhang, J., Liu, H., Liu, X.-S., and Shi, H.-B. (2015). The histogram analysis of diffusion-weighted intravoxel incoherent motion (IVIM) imaging for differentiating the Gleason grade of prostate cancer. European Radiology, 25, 994–1004.
20. Gašparović, B., Mauša, G., Rukavina, J., and Lerga, J. (2023). Evaluating Yolov5, Yolov6, Yolov7, and Yolov8 in underwater environment: Is there real improvement? In Evaluating Yolov5, Yolov6, Yolov7, and Yolov8 in Underwater Environment (pp. 1–4). IEEE.
21. Flach, P., and Kull, M. (2015). Precision-recall-gain curves: PR analysis done right. Advances in Neural Information Processing Systems, 28.
22. Krstinić, D., Braović, M., Šerić, L., and Božić-Štulić, D. (2020). Multi-label classifier performance evaluation with confusion matrix. Computer Science & Information Technology, 1, 1–14.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 American Journal of Psychiatric Rehabilitation
This work is licensed under a Creative Commons Attribution 4.0 International License.
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License permitting all use, distribution, and reproduction in any medium, provided the work is properly cited.
This work is licensed under a 
