A Method for Large Underground Structures Geometry Evaluation Based on Multivariate Parameterization and Multidimensional Analysis of Point Cloud Data

In underground mining, the tunnels are performed mainly by blasting. Unfortunately, their quality may be different than expected. It is important to monitor both the quality and changes of existing tunnels. The length of corridors in underground mines is massive. Even if one wants to limit monitoring to those used every day for main technological processes it is a vast amount of work. Any stationary monitoring system cannot be utilized both due to everyday blasting procedures and mobile machines operation. The method proposed here is based on LiDAR/TLS measurements to obtain a point clouds. Data processing procedures are proposed to extract parameters describing the geometry of the tunnel. Statistical parameters are selected to describe cross-sections for final 1D features analysis along the tunnel length. Such vector may serve as a basis for blasting evaluation. The methodology was tested and validated for the data obtained in historical mine in Poland