激光SLAM技术在巷道精细建模的应用研究  

Research on Application of Laser SLAM Technology in Fine Modeling of Roadway

作　　者：刘敬东 李旭 于凤启 苟丙荣 贺国庆 巩泽文 Liu Jingdong;Li Xu;Yu Fengqi;Gou Bingrong;He Guoqing;Gong Zewen(Guodian Construction Investment Inner Mongolia Energy Co.,Ltd.,Ordos 017209,China;CCTEG Xi’an Research Institute(Group)Co.,Ltd.,Xi’an 710054,China)

机构地区：[1]国电建投内蒙古能源有限公司,内蒙古鄂尔多斯017209 [2]中煤科工西安研究院(集团)有限公司,西安710054

出　　处：《煤矿机械》2024年第10期199-202,共4页Coal Mine Machinery

基　　金：国家能源集团重点研发项目(GJNY-21-154);中煤科工集团面上项目(2023XAYKF09)。

摘　　要：为了解决井下数据探测成本高、难度大和测量盲区多等问题,提出一种基于激光实时定位与地图构建(SLAM)技术的井下巷道三维建模方法。通过对点云数据进行坐标转换、点云配准和纹理映射等方法实现了陕北某煤矿巷道井下精细建模。结果显示,构建的点云精细模型坐标转换平均误差为24.519 mm,纹理映射平均误差为5.23像素、最大误差为12.87像素,三维模型坐标平均误差10.45 mm。这表明基于激光SLAM技术的巷道建模技术满足井下测量要求,能够为煤矿安全生产提供详实的巷道数据。In order to solve the problems of high cost,high difficulty and many blind zones of underground data detection,proposed a three-dimensional modeling method of underground roadway based on laser simultaneous localization and mapping(SLAM)technology.Through the coordinate conversion of point cloud data,point cloud alignment and texture mapping methods,achieved a coal mine roadway underground fine modeling in northern Shaanxi province.The results show that the average error of coordinate conversion of the constructed point cloud fine model is 24.519 mm,the average error of texture mapping is 5.23 pixels,the maximum error is 12.87 pixels,and the average error of coordinates of the 3D model is 10.45 mm,which suggests that the roadway modeling based on the laser SLAM technology meets the requirements of the underground measurements and can provide detailed data of the roadway for the safety production of coal mine.

关 键 词：激光SLAM 点云拼接 点云配准 纹理贴图 精细建模 

分 类 号：P168.13[天文地球—天文学]