Research on the Realization Method of 3D Model Visualization of Mine Roadways
The digitization and visualization of roadways have been a research hotspot in the mining field in recent years. To seek an efficient and convenient method for roadway visualization,this paper starts with image sets obtained from photogrammetry,uses the Structure from Motion (SfM) algorithm combined with the multi-view stereo matching algorithm based on patches to reconstruct the surface point cloud model of the roadway. After mesh division and texture mapping of the surface point cloud model,a 3D real-scene model of the roadway is obtained. Taking the roadway of No. 3 mine shaft in Changle mining area,Shaoyang,Hunan as the research object,a roadway model with a length of about 15 m is reconstructed. The control point distance is selected as the model accuracy discrimination index,and it is found that the maximum error between the measured control point distance and the control point distance in the 3D real-scene model does not exceed 1.5 cm. In addition,comparing the displacement of roadway monitoring points at different time periods with the displacement obtained from the model,it is found that the relative error is controlled within 4.4％. This verifies the feasibility of the SfM algorithm in the research of mine roadway visualization. The generated 3D real-scene model of the roadway has rich details and realistic textures. Monitoring deformation based on the 3D real-scene model of the roadway meets the accuracy requirements and has broad prospects in underground remote operations.
RoadwayVisualizationStructure from Motion (SfM)Photogrammetry
NETL
NSTL
万方数据
