A Deep Learning Approach to Identify Rock Bolts in Complex 3D Point Clouds of Underground Mines Captured Using Mobile Laser Scanners

2025-06-25

AI Summary

Automated detection of rock bolts—small, often occluded by shotcrete, and highly susceptible to noise and complex backgrounds—in large-scale 3D point clouds acquired via mobile laser scanning in underground mines remains challenging. To address this, we propose DeepBolt, a two-stage deep learning architecture specifically designed for extreme class imbalance. DeepBolt integrates 3D point cloud semantic segmentation with an adaptive feature enhancement mechanism to significantly improve robustness in detecting small objects. Evaluated on a real-world mine dataset, DeepBolt achieves 96.41% precision and 96.96% recall, with a mean Average Precision (mAP) that surpasses the state-of-the-art by 42.5%. These results demonstrate its effectiveness and reliability under low-light, high-clutter underground conditions. DeepBolt thus provides a practical, deployable solution for intelligent safety monitoring in mining operations.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Intelligent Robots: State Estimation
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Security and Privacy: Large-scale security measurements
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Rock bolts are crucial components of the subterranean support systems in underground mines that provide adequate structural reinforcement to the rock mass to prevent unforeseen hazards like rockfalls. This makes frequent assessments of such bolts critical for maintaining rock mass stability and minimising risks in underground mining operations. Where manual surveying of rock bolts is challenging due to the low light conditions in the underground mines and the time-intensive nature of the process, automated detection of rock bolts serves as a plausible solution. To that end, this study focuses on the automatic identification of rock bolts within medium to large-scale 3D point clouds obtained from underground mines using mobile laser scanners. Existing techniques for automated rock bolt identification primarily rely on feature engineering and traditional machine learning approaches. However, such techniques lack robustness as these point clouds present several challenges due to data noise, varying environments, and complex surrounding structures. Moreover, the target rock bolts are extremely small objects within large-scale point clouds and are often partially obscured due to the application of reinforcement shotcrete. Addressing these challenges, this paper proposes an approach termed DeepBolt, which employs a novel two-stage deep learning architecture specifically designed for handling severe class imbalance for the automatic and efficient identification of rock bolts in complex 3D point clouds. The proposed method surpasses state-of-the-art semantic segmentation models by up to 42.5% in Intersection over Union (IoU) for rock bolt points. Additionally, it outperforms existing rock bolt identification techniques, achieving a 96.41% precision and 96.96% recall in classifying rock bolts, demonstrating its robustness and effectiveness in complex underground environments.

Problem

• Automate rock bolt detection in 3D mine point clouds
• Overcome noise and occlusion in large-scale laser scans
• Improve accuracy over traditional feature-based methods

Innovation

• DeepBolt two-stage deep learning architecture
• Handles severe class imbalance effectively
• Outperforms state-of-the-art semantic segmentation models