Research on Deformation Detection Technology of Coal Mine Underground Roadway Based on Feature Point Segmentation

The deformation detection of coal mine underground tunnels is one of the vital safeguards for safe production. However, due to the massive scale of the tunnels and the drawback of weak texture features, there has been no highly precise and easy-to-operate detection solution to date. This paper studies a deformation detection technology for coal mine underground roadways based on feature point segmentation, aiming to achieve efficient deformation detection in coal mine environments. By placing circular retroreflectors within the scene, the LiDAR mapping data is segmented. Subsequently, after a second scan and mapping, the deformation of each section of the tunnel model is calculated through ICP point cloud registration. To address the weak texture features of underground roadways, the deployment of reflective panels not only enhances feature points but also provides precise positioning information. Segmented registration detection effectively avoids the cumulative errors in LiDAR mapping, with the deformation detection accuracy reaching within 3 cm.