Parallel Progressive Radiosity with Adaptive Meshing

Progressive radiosity is widely used for realistic image synthesis in computergraphics applications. High-quality image generation usually requiresradiosity with adaptive patch refinement to account for global illuminationeffects from irregular objects in a 3D space. Parallelizing such an algorithmwith adaptive refinement is difficult since computation cost for each objectvaries from one iteration to another depending on the location of dynamicallyselected shooting patches. Dynamic load balancing is required but its overheadis high for distributed memory systems. This paper presents an efficientparallel algorithm for progressive radiosity, which adopts a static processorassignment strategy to take advantages of hierarchical computation structure inthis problem, minimize communication and balance dynamic load. Our experimentson a Meiko CS-2 distributed memory machine show that this algorithm hasachieved good performance for the tested cases.