Memory Management Schemes for Radiosity Computation in Complex Environments

Hierarchical radiosity is a very demanding process in terms of computation time and memory ressources even for scenes of moderate complexity. To handle complex environments, not fitting in memory, new solutions have to be devised. One solution is to partition the scene into subsets of polygons (3D cells or clusters) and to maintain in memory only some of them. The radiosity computation is performed only for this resident subset which changes during the resolution process. This change entails many read and write operations from or onto the disk. These disk transfers must be ordered to make the radiosity algorithms tractable. We propose in this paper different ordering strategies which can be seen as complementary to those devised by Teller [1].