3D Global Localization in the Underground Mine Environment Using Mobile LiDAR Mapping and Point Cloud Registration.

Area of Science:

• Geospatial Engineering
• Robotics
• Mining Technology

Background:

• Accurate 3D mapping and localization are critical for underground mining operations.
• Existing methods often struggle with the dynamic and complex environments found in mines.
• Mobile LiDAR technology offers a promising solution for detailed 3D data acquisition.

Purpose of the Study:

• To develop and validate a 3D global localization method for underground mines.
• To enable precise recognition and tracking of object locations using mobile LiDAR.
• To assess the accuracy and robustness of the proposed localization technique.

Main Methods:

• Mobile LiDAR mapping to create an initial global point cloud map of the mine.
• Generation of local LiDAR scans for specific underground positions.
• Feature extraction using Fast Point Feature Histogram (FPFH) descriptors.
• Point cloud registration via Random Sample Consensus (RANSAC) and Iterative Closest Point (ICP) algorithms.
• Localization using the LiDAR sensor trajectory within the global coordinate system.

Main Results:

• Successful alignment of local point cloud datasets with the global map in six underground mine areas.
• Achieved a low average localization error of approximately 0.13 m.
• Demonstrated consistent performance across different mobile LiDAR systems.
• Validated the ability to track changes in dynamic object positions over time.

Conclusions:

• The proposed 3D global localization method effectively utilizes mobile LiDAR and point cloud registration for accurate underground mine mapping.
• The technique provides reliable and precise localization, crucial for operational safety and efficiency.
• The study confirms the viability of mobile LiDAR for real-time spatial awareness in subterranean environments.