Multi-Resolution Modelling of the Underground Mining Workplace by Integrating Laser Scanning and Close-Range Photogrammetry

Abstract

This research highlights the essential requirement for precise deformation monitoring in underground mining, where conventional measurement techniques frequently compromise mine personnel safety. The main aim of this research is to improve spatial accuracy in deformation monitoring by integrating terrestrial laser scanning (LiDAR) and close-range photogrammetry (CRP) datasets. Data was acquired by both LiDAR and CRP technologies, and the data was integrated to produce a more accurate depiction of the underground mines. To assess the precision of the fused model, measurements derived from it were assessed using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Percentage Error as the principal evaluation metrics. The fused 3D model displayed the lowest MAE and RMSE values relative to individual LiDAR and CRP data, indicating enhanced accuracy and diminished error margins. The integrated model attained a Mean Absolute Error (MAE) of 0.021 m (~2.1 cm) and a Root Mean Square Error (RMSE) of 0.027 m (~2.7 cm), demonstrating considerable accuracy compared to models from their respective sensors. In the context of underground deformation monitoring, where measurement tolerance of 5 cm or greater are often acceptable, these error magnitudes indicate high accuracy.