EventsOHOW 2026 – The 5th International Symposium on One Health, One World
Published
This submission belongs to event OHOW 2026 – The 5th International Symposium on One Health, One World
Published date
18 Nov, 2026
Contribution&Funding
Conceptualization,
[Zhichen Wang]
Methodology,
[Zhichen Wang]
Validation,
[Zhichen Wang]
Formal analysis,
[Zhichen Wang]
Investigation,
[Zhichen Wang]
Resources,
[Tsukasa Mizutani]
Data curation,
[Tsukasa Mizutani]
Writing - original draft,
[Zhichen Wang]
Writing - review,
[editing] ; [Tsukasa Mizutani]
Visualization,
[Zhichen Wang]
Supervision,
[Tsukasa Mizutani]
Project administration,
[Tsukasa Mizutani]
Funding acquisition,
[Tsukasa Mizutani]
Funding: --
Citation
Zhichen Wang, Tsukasa Mizutani, An Uncertainty-Aware Framework for Underground Cavity Detection and Segmentation in Real-World 3D GPR Images Using  Ensemble Learning, in Proceedings of OHOW 2026 – The 5th International Symposium on One Health, One World, Hanoi, 26 November–28 November 2026, MDPI: Basel, Switzerland
Share
Email
Facebook
Twitter
LinkedIn

An Uncertainty-Aware Framework for Underground Cavity Detection and Segmentation in Real-World 3D GPR Images Using  Ensemble Learning

1. Institute of Industrial Science, University of Tokyo, Tokyo, 153-8505, Japan
Abstract

Automated interpretation of ground penetrating radar (GPR) images is essential for efficient underground cavity detection and segmentation. However, existing methods often exhibit limited performance in real-world applications due to the high variability in cavity appearances, which poses significant challenges to the generalization and reliability of deep learning models. To address this issue, this study proposes an uncertainty-aware framework using ensemble learning for underground cavity detection and segmentation in real-world 3D GPR images. The proposed framework ensembles multiple segmentation models trained with different random initializations to capture variations in model predictions and improve robustness against diverse cavity appearances. By integrating predictions from these models, the framework enhances segmentation performance while enabling the quantification of predictive uncertainty through inter-model disagreement. An uncertainty-based filtering strategy is further introduced to identify and filter out unreliable cavity detections, thereby reducing false positives and improving detection reliability. The proposed framework is validated using 3D GPR images collected from real road networks, demonstrating its effectiveness in improving cavity detection and segmentation under realistic inspection conditions. The results highlight the benefits of combining ensemble learning with uncertainty quantification to address the challenges posed by highly variable cavity appearances. This study provides an uncertainty-aware approach to automated GPR interpretation, contributing to more robust and reliable underground cavity detection and supporting the practical application of GPR technology in subsurface infrastructure inspection.

Keywords
Ground Penetrating Radar
Object Detection
Underground Cavities
Uncertainty Quantification
Development of Screen-Printed Disposable 3-in-1 Electrodes for On-Site Heavy Metals Analyzers
Numerical Investigation of Hidden Asphalt Pavement Cracks Using Ground-Penetrating Radar