A Multiple-Scale Stochastic Modelling Primitive

AbstractStochastic modeling has been successfully used in computer graphics to model a wide array of natural phenomena. In modeling three-dimensional fuzzy or partially translucent phenomena, however, many approaches are hampered by high memory and computation requirements, and by a general lack of user control. We will present a general stochastic modeling primitive that operates on two or more scales of visual detail, and which offers considerable flexibility and control of the model. At the macroscopic level, the general shape of the model is constrained by an elliptical correlation function that controls the interpolation of user-supplied data values. We use a technique called Kriging to perform the interpolation. The microscopic levle permits the addition of noise, which allows a user to add interesting visual textural detail and translucency. A wide variety of noise-synthesis techniques can be employed in our model. We shall describe the mathematical structure of our model, and give an attractive rendering implementation that can be embedded in a traditional ray tracer rather than requiring a volume renderer. As an example, we shall apply our approach to the modeling of clouds.