机电工程技术2025，Vol.54Issue(11)：175-180,6.DOI:10.3969/j.issn.1009-9492.2025.11.033
基于三维点云分割组合的煤堆三维重构和体积测量算法
Research on 3D Reconstruction and Volume Measurement Algorithm for Coal Pile Based on 3D Point Cloud Segmentation and Combination
孙佳胜 1朱新理 1罗强 2王琰 3李根生 4杨浩基1
作者信息
- 1. 徐州意泰林电子科技有限公司,江苏 徐州 221000
- 2. 河南大有能源股份有限公司新义煤矿,河南 洛阳 471842
- 3. 河南大有能源股份有限公司新安煤矿,河南 洛阳 471842
- 4. 河南科技大学车辆与交通工程学院,河南 洛阳 471000
- 折叠
摘要
Abstract
The coal storage yards in large coal mines have numerous coal piles with complex surface structures and irregular shapes.Traditional manual measurement methods encounter difficulties in implementation and exhibit low accuracy.To solve the above-mentioned problems,a 3D reconstruction and volume measurement algorithm of coal pile based on 3D point cloud segmentation and combination is proposed.Additionally,a method for measuring the volume of coal piles using point cloud data is introduced.Firstly,3D point cloud data of the irregular coal pile surface is obtained through laser radar scanning.Then,reinforcement learning algorithm is employed to optimize the Delaunay triangulation network division algorithm,and the point cloud data is effectively segmented and combined to achieve accurate 3D modeling of the coal pile.Finally,the volume of the coal piles is calculated using the height information from the 3D point clouds.The method is tested and verified with simulated coal pile point cloud data.The results demonstrate that the proposed method can use reinforcement learning theory to improve the quality of Delaunay triangulation network division,and significantly improve the accuracy and reliability of the 3D reconstruction effect and volume measurement of coal piles.Moreover,the result of coal pile volume measurement only has a deviation of 1.27%from the actual value,which meets the accuracy requirements of coal mines for coal pile volume measurement.It provides strong technical support for the management and planning of coal storage yards in coal mines.关键词
体积测量/点云分割/强化学习/三维重构/煤堆Key words
volume measurement/point cloud segmentation/reinforcement learning/3D reconstruction/coal piles分类
计算机与自动化引用本文复制引用
孙佳胜,朱新理,罗强,王琰,李根生,杨浩基..基于三维点云分割组合的煤堆三维重构和体积测量算法[J].机电工程技术,2025,54(11):175-180,6.
