-
Notifications
You must be signed in to change notification settings - Fork 0
Planet Paper
| Publications / Theses → Planet Paper (IEEE 2020) |
|---|
Dillon Fisher, Michael McCarthy, Connor Jakubik, Patrick Zhong, Wade Smonko, Kendall Mares, and Gregory Chamitoff
Procedurally generated terrain is widely used in the gaming industry, and has seen increased use for engineering VR simulations in recent years. These terrain generation methods allow for rapid development of high-fidelity environments, and since the generation is seeded in coherent noise functions, require very little storage space. A simulation requiring high-fidelity height maps for known celestial bodies like the Moon or Mars, cannot solely use procedural terrain. This paper presents a system for generating surface meshes through the combination of a Digital Elevation Model (DEM) and procedural terrain techniques, which allows a planet to have scientifically defined macroscopic detail, while leaving finer levels of detail (LODs) to be determined by Perlin noise functions. This provides a recognizable topography for a rendered planet as viewed from orbit, while also retaining realistic, procedurally generated terrain on the surface. Using a quadtree data structure, we can optimize the rendering of the planet at multiple levels of detail, dynamically displaying higher levels of detail around the user and minimal detail at points of the surface far from the user. Typical planetary DEMs do not have high surface spatial resolution. The Lunar Orbiter Laser Altimeter (LOLA), for example, has collected height data from the Lunar surface, generating a full DEM with a surface spatial resolution of 118 meters. The system presented here is designed to display measured height data to the user until it reaches a level of detail beyond the original DEM. At this point, the system interpolates between those data points with procedural terrain. This system has been used to render most of the inner planets in our solar system and the Moon at their proper scales using the DEMs provided from the United States Geological Survey (USGS). The system is lightweight enough to be used in a larger VR simulation, with both surface and orbital operations.