An Identification Tool to build Physical Models for Virtual Reality

Abstract

Particle-based physical models appear to be a promising general framework to build realistic and efficient models for simulation and animated image synthesis. They are being used increasingly to model smokes elastic or articulated bodies, fabrics, but they need to be associated to identification methods to deserve the “physical model” qualification and to ensure their behavioral realism. The evolutionary parameter identification technique proposed in this chapter has proven successful to reconstruct model internal parameters from their kinematic outputs—including non-linear, conditional, elastic, and viscous bonds in periodic or non-periodic objects—and thus provides the particle-based modeling tool with the “measuring instrument” that should always be associated to a physical model.