Dynamics Modeling and Culling

Procedural animation, via dynamical systems, has many advantages over keyframe animation, yet suffers from high computational cost and difficulties in modeling. We describe tools that analyze and process systems to enable culling if the system is not in view, based on a qualitative model of viewer prediction. Application of these tools results in the cost of dynamic simulation growing linearly with the number of systems in view, regardless of the number of systems out of view. To address the difficulties of modeling, we present a runtime layer that interfaces between a VRML browser and simple descriptions of dynamics written in Java. To create these runtime environments, we developed a modeling program in which a user interactively builds models and previews their dynamics. An important outcome of this work is the emergence of a library of dynamical systems, each of which is simple to incorporate into a virtual world. Together, our tools enable large numbers of complex dynamic models to be efficiently and easily included in a VRML world while maintaining high frame rates.