An Efficient Radiosity Approach for Realistic Image Synthesis

The radiosity method models the interaction of light between diffusely reflecting surfaces and accurately predicts the global illumination effects. Procedures are now available to simulate complex environments including occluded and textured surfaces. For accurate rendering, the environment must be discretized into a fine mesh, particularly in areas of high intensity gradients. The interdependence between surfaces implies solution techniques which are computationally intractable. This article describes new procedures to predict the global illumination function without excessive computational expense. Statistics indicate the enormous potential of this approach for realistic image synthesis, particularly for dynamic images of static environments.