Optimizing artificial lighting conditions for rock plane detection on tunnel faces: A controlled experiment using smartphone videogrammetry

The study investigates the influence of artificial lighting conditions on the accuracy of detecting rock planes on tunnel faces using videogrammetry. A controlled experiment was conducted employing 3D-printed models with predefined dip angles and dip directions, illuminated under varying conditions of light intensity, color temperature, and angles. The research utilized advanced imaging techniques to capture video data at multiple distances and angles, followed by detailed 3D point cloud reconstruction and analysis. Results demonstrate that lighting parameters significantly affect videogrammetric outcomes. Optimal configurations such as 5500 K color temperature and 1000 lx intensity enhanced point cloud density and reduced errors, achieving precise detection of rock planes. Oblique video captures outperformed straight-on captures, providing better coverage of inclined surfaces and reducing Root Mean Square Error (RMSE) in measurements. In contrast, combined lighting setups introduced inconsistencies, leading to increased variability. The findings underline the importance of tailored lighting and capture strategies to improve detection accuracy in tunnel environments. Practical guidelines derived from this research offer valuable insights for optimizing videogrammetric workflows in engineering and geotechnical applications.