Deep learning-based object detection on LiDAR-derived hillshade images: insights into grain size distribution and longitudinal sorting of debris flows

Schmid, Paul E.; Hirschberg, Jacob; Spielmann, Raffaele; Aaron, Jordan

doi:10.5194/nhess-25-4863-2025

Articles | Volume 25, issue 12

https://doi.org/10.5194/nhess-25-4863-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/nhess-25-4863-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 25, issue 12

Research article

|

09 Dec 2025

Research article |

| 09 Dec 2025

Deep learning-based object detection on LiDAR-derived hillshade images: insights into grain size distribution and longitudinal sorting of debris flows

Paul E. Schmid, Jacob Hirschberg, Raffaele Spielmann, and Jordan Aaron

Data sets

Deep learning-based object detection on LiDAR-derived hillshade images: Insights into grain size distribution and longitudinal sorting of debris flows Paul Schmid et al. https://doi.org/10.3929/ethz-c-000787515

Data for Detailed observations reveal the genesis and dynamics of destructive debris-flow surges Jordan Aaron et al. https://doi.org/10.3929/ethz-b-000736836

Short summary

Debris flows are fast-moving water-sediment mixtures in steep channels, posing risks to infrastructure and lives. Traditional analysis is slow and labor-intensive. This study presents a method using laserscanners and deep learning to detect and track moving objects during active events. By converting three-dimensional data to two-dimensional images, it enables fast, accurate measurement of object speed and size. This improves debris-flow monitoring, enhancing hazard understanding and mitigation.