Modeling and rendering of gaseous phenomena using particle maps

Abstract

Any modeling scheme for gaseous phenomena in graphics has to capture three aspects: the fuzzy geometry of the gas, the dynamics, characterized by the presence of vortices, and the interaction of light with the gaseous volume. We represent the gaseous volume as a particle system and apply Vortex Element Methods (VEM) to model the dynamics. A Lagrangian formulation that is gridless and hence ideal for unbounded flows is used. A gridless approach to ray tracing the particle systems is developed using particle maps. These maps are used to estimate densities within a gaseous volume analogous to the way volume photon maps are used to estimate radiance during Monte Carlo ray tracing. A technique is proposed to merge particle and volume photon maps to obtain an effective method for simulating multiple scattering in a dynamic inhomogeneous participating medium. Our method for modeling and rendering gaseous phenomena is conceptually simple and grid free. Particle maps play an effective role, as the nearest neighbor information obtained during the rendering phase is exploited during the dynamics computation. We preset results that demonstrate the effectiveness of our approach. Copyright © 2000 John Wiley & Sons, Ltd.

A survey on participating media rendering techniques

“…Stochastic methods solve the transport equation by means of random sampling, using random paths along interaction points. We distinguish between the methods that set the interaction points by using a constant step distance [9,10], from those that sample a function of κ t , which include light tracing [66,68], bidirectional path-tracing [46], photon maps [37,47,23,1] and Metropolis Light Transport [67]. Another categorization is made according to the view dependency of the methods.…”

“…Adabala and Manohar [1] use particle systems to model gaseous volumes, allowing the treatment of inhomogeneous anisotropically scattering media. By displacing the particles following fluids dynamics equations, the model can evolve in time, avoiding the use of grids.…”

Techniques for Realistic Visualization of Fluids: A Survey

“…The work which uses the Lagrangian formulation for dynamics simulation include the scientific visualization paper of Ma et al [36] where a simulation of turbulent mixing of gases and particles is visualized. Our work [37] which uses Vortex Element Methods (VEM) [33] techniques to simulate gaseous dynamics also belongs to this class.…”

“…A graphics tools developer can use this framework to evolve new techniques to visualize fluids by exploring combinations of methods not explored before. As an example, the technique of particle maps [37], has been developed by the authors as a method that combines the benefits of texture and particle based methods.…”

Global illumination techniques for the computation of hight quality images in general environments

“…Stochastic methods solve the transport equation by means of random sampling, using random paths along interaction points. We distinguish between the methods that set the interaction points by using a constant step distance [20,21], from those that sample a function of κ t , which include light tracing [114], bidirectional path-tracing [83], photon maps [71,84,44,7] and Metropolis Light Transport [116]. Another categorization is made according to the view dependency of the methods.…”