Indoor Coal Inventory Algorithm Using UAV-Based LiDAR SLAM

This paper proposes an indoor coal inventory method utilizing UAV-mounted laser Simultaneous Localization and Mapping (SLAM) technology to address the limitations of traditional measurement approaches in accuracy, efficiency, and safety. By tightly coupling LiDAR data with IMU on a UAV platform, our system achieves comprehensive 3D data acquisition in complex indoor environments. The key innovations include: (i) a RANSAC-based trajectory plane fitting and point cloud rotation method that enhances data alignment robustness by reducing noise impact, (ii) an automated ground segmentation approach combining cloth simulation filtering (CSF) with Euclidean clustering for precise coal pile extraction, and (iii) a quadric surface parameter model for ground deformation compensation to improve volume calculation accuracy. Experimental results demonstrate volume measurement errors of only 2.19% and 0.6% for two typical coal piles, significantly outperforming conventional methods. This work provides an automated, high-precision solution for coal inventory management, advancing intelligent stockpile monitoring.