Rock and Debris Fall Detection Using Total Gray Level Method

Sudhan Regmi; Ko-Fei Liu; Ranjan Kumar Dahal

Authors

Sudhan Regmi Department of Civil Engineering, National Taiwan University, Department of Civil Engineering, National Taiwan University, Taipei, Taiwan Author
Ko-Fei Liu Department of Civil Engineering, National Taiwan University, Department of Civil Engineering, National Taiwan University, Taipei, Taiwan Author
Ranjan Kumar Dahal Central Department of Geology, Tribhuvan University, 3Central Department of Geology, Tribhuvan University, Kritipur, Kathmandu, Nepal Author

Keywords:

Debris fall detection, Early warning, Image analysis, Rockfall detection, Total gray model

Abstract

This study investigates the application of the total gray level method for identifying rock and debris falls through video analysis, offering a viable alternative to resource-heavy machine learning techniques. The method focuses on variations in total grayscale intensity within a specified Region of Interest (ROI) and establishes a detection threshold informed by environmental noise levels. Initial tests showed that this approach effectively detected ongoing rock and debris falls while requiring minimal computational resources. The findings indicated that while a threshold set at twice the noise level was too sensitive, increasing it to five times the noise level considerably enhanced accuracy.

Author Biographies

Ko-Fei Liu, Department of Civil Engineering, National Taiwan University, Department of Civil Engineering, National Taiwan University, Taipei, Taiwan

Professor
Ranjan Kumar Dahal, Central Department of Geology, Tribhuvan University, 3Central Department of Geology, Tribhuvan University, Kritipur, Kathmandu, Nepal

Associate Professor

References

Liu K.-F., Kuo T.-I. and Wei, S.-C. (2021). Debris Flow Detection Using a Video Camera. In N. Casagli, V. Tofani, K. Sassa, P. T. Bobrowsky, and Takara K. (Eds.), Understanding and Reducing Landslide Disaster Risk: Volume 3 Monitoring and Early Warning (pp. 141-147).

Noël F., Jaboyedoff M., Caviezel A., Hibert C., Bourrier F. and Malet J. P. (2022). Rockfall trajectory reconstruction: a flexible method utilizing video footage and high-resolution terrain models. Earth Surf. Dynam., 10(6), 1141-1164.

Pham M.-V. and Kim Y.-T. (2022). Debris flow detection and velocity estimation using deep convolutional neural network and image processing. Landslides, 19(10), 2473-2488.