For the complete documentation index, see llms.txt. This page is also available as Markdown.

6905 - INVESTIGATING DATA COLLECTION AND PROCESSING STRATEGIES FOR UAV-BASED TERRAIN MODELS IN VEHICLE DYNAMICS SIMULATION

Paper presented at ISTVS 2025 | 55th Conference of the International Society for Terrain-Vehicle Systems https://doi.org/10.56884/2I9ZL94A

Authors: Herman Hamersma, Christian Van Aswegen, Glenn Guthrie, Schalk Els, Carl Becker

Keywords: Terrain modelling; road profiling; off-road mobility

Abstract:

Accurate modelling of the interaction between a tyre and the terrain is essential for effective vehicle dynamics simulation, especially in challenging off-road conditions. A previous study outlined the development of three-dimensional terrain models using images captured by an unmanned aerial vehicle (UAV) and suggested several avenues for further investigation. This study addresses those recommendations, which include: (1) examining the impact of various image collection strategies, such as different image overlap ratios, varying flight path altitudes, and using both orthographic and oblique images, and (2) expanding the previous results to encompass a broader range of terrains, including a comparison of experimental measurements from a test vehicle with simulation results. Additionally, the interpolation technique used to regularise the point cloud data was found to significantly affect the terrain model. Images were collected across multiple test tracks at a proving ground for which baseline measurements from a traditional mechanical profilometer were available for direct comparison with the UAV-obtained data. The findings reveal that while constructing three-dimensional terrain models from UAV data is straightforward, several potential pitfalls must be carefully avoided, as they can significantly affect the accuracy of the results. The results provide a step-by-step guide to developing terrain models of various surfaces, from relatively smooth man-made roads (e.g., an ISO 8608:2016 Class-C road) to very rough natural terrain, opening various avenues for future mobility research in challenging terrains.

Last updated