A Culling and Prediction Dynamics Modeling and Culling Ieee Computer Graphics and Applications

nimating geometry procedurally, using a dynamical system (which consists of a set of state variables and a set of equations for describing how those variables change over time) rather than with keyframes, offers the advantages of physical realism , interactivity, compact descriptions, and infinite variety of motion. However, the use of dynamical systems has been inhibited by two factors: their computational cost, even when out of view, and the difficulties of implementation, partially due to a lack of standard modeling and runtime environments. VRML is particularly prone to these problems. The most suitable, currently available, scripting language is Java. Controlling a VRML world from a Java object requires knowledge of a large number of methods and types, and is highly error prone. A lack of debugging environments, or even consistent exception handling, only exacerbates the problem. Our research aims to develop modeling tools that enable incorporating large numbers of efficient dynam-ical systems into virtual environments, while abstracting the modeling process as much as possible. To achieve efficiency, we concentrate on culling dynamical systems: if the system is not in view, we do not compute any dynamics for it. As a concrete example of the benefits of culling, consider a museum of kinetic sculpture (such as mobiles). Such a world has a large number of dynamic objects— the exhibits—placed in mutually nonvisible rooms. Traditional dynamic simulation requires computing the motion of all the sculptures on every frame, even if some sculptures aren't visible to the viewer. Naturally, this would severely limit the size of the museum. Culling avoids dynamic state computation for the exhibits not in view, just as geometry culling avoids rendering their geometry. Exploiting culling allows the museum to be arbitrarily large, provided only a few rooms are visible at once. Similar examples occur in factory simulations or dense city environments. In this article we describe three tools that together provide an environment for authoring cullable, dynamic , rigid-body objects in VRML and Java: s A code transformation tool that exploits approximations to dynamical systems to enable culling. The frame rate for a world using these systems depends only on the number of systems in view. This results in significant speedups over conventional models and enables very large virtual worlds. s A runtime layer generator, which defines a simple standard interface between a VRML browser and dynamical systems described in Java. This hides the complexities of the underlying interface …