Abstract
Using dynamical systems rather than keyframing to animate a world is a desirable yet computationally expensive approach. We present techniques for culling dynamical systems that avoid unnecessary computation, and describe tools for automating much of the required work. Based on qualitative observations of how viewer's predict dynamical state over time, we identify methods for generating state while ensuring consistency, which we define as ensuring that a viewer's predictions are satisfied. Our tools take as input a description of a dynamical system, and produce an alternate description that may be efficiently culled. We also describe an interactive modeler in which authors attach dynamic variables to geometric transformations, allowing the modular re-use of dynamical systems. 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.
Original language | English (US) |
---|---|
Pages | 15-24 |
Number of pages | 10 |
DOIs | |
State | Published - 1998 |
Externally published | Yes |
Event | Proceedings of the 1998 3rd Symposium on the Virtual Reality Modeling Language, VRML'98 - Monterey, CA, USA Duration: Feb 16 1998 → Feb 19 1998 |
Other
Other | Proceedings of the 1998 3rd Symposium on the Virtual Reality Modeling Language, VRML'98 |
---|---|
City | Monterey, CA, USA |
Period | 2/16/98 → 2/19/98 |
ASJC Scopus subject areas
- Modeling and Simulation