Weathering and debris simulation on high-resolution voxel scenes

Weathering and debris simulation on high-resolution voxel scenes

Kuijpers, Mika (TU Delft Electrical Engineering, Mathematics and Computer Science; TU Delft Computer Graphics and Visualisation)

Eisemann, E. (mentor)
Dijkstra, Y.M. (graduation committee)
Billeter, M.J. (graduation committee)

Delft University of Technology

Computer Science

2023-06-14

Voxels are cells on a 3D regular grid. Voxel-based scenes have many applications, including frequent use in simulations or games. Over the years, the field of high-resolution voxel scenes has progressed significantly, allowing for compressing and real-time editing of high-resolution scenes. This possibility of editing high-resolution scenes led to various editing tools. This thesis aims to expand this set of tools with a focus on more complex simulation-based solutions. We explored the field of terrain weathering and decided on a spheroidal weathering tool to perform weathering on voxel-based scenes. One of the existing limitations of this particular method was that it did not account for debris simulation. We ultimately decided to add a granular simulation using a layer-based approach that integrates with the weathering tool but is also able to function in a standalone manner.
This thesis presents three highly-customizable editing tools for voxel-based scenes: spheroidal weathering, granular simulation, and a combined tool that enables weathering with debris simulation. These tools are integrated into the HashDAG framework, which is a high-resolution voxel-grid method building upon a directed acyclic graph representation of a sparse voxel octree. Our solutions enable close to real-time editing for changes with a smaller regional support. There is room for improvement in terms of performance at scale, with various potential ideas presented in the future work section.

voxel
weathering
spheroidal weathering
granular simulation
debris simulation
simulation
CUDA

http://resolver.tudelft.nl/uuid:38603eac-e9c1-4814-81e2-73aa2df9c25a

Student theses

master thesis

© 2023 Mika Kuijpers