Terrain Rendering Using Geometry Clipmaps
Nick Brettell
Department of Computer Science
University of Canterbury
Abstract
A primary difficulty in terrain rendering is displaying realistic terrains to the user at real-time frame rates. The brute force approach is usually too complex for real-time frame rates to be achieved. Several terrain-rendering techniques have been proposed that use Level of Detail (LOD) to generate a simplified representation of a terrain. The geometry clipmap is a recently proposed approach that utilises the potential of modern graphics hardware. It stores vertex data on the graphics card, that is updated incrementally as the viewpoint moves. LOD is achieved using regular nested grids of increasing size and decreasing detail, centred around the viewpoint. We implemented the geometry clipmaps algorithm, and present aspects of our implementation, such as the maximum number of clipmap levels that are possible, and constraints on the extent of each level. We also performed a comparative analysis with other terrainrendering techniques. Geomipmapping outperformed geometry clipmaps on a midrange graphics card, but geometry clipmaps had the greatest rendering throughput on high-end graphics hardware.
