The Population-Based Optimization Algorithms for Role Modelling and Path Generation in Group Animation

摘要

Traditional animation by key frame techniques takes animators lots of time and vigour to model and simulate behaviours of crowds. For solving this problem, this paper presents a novel group animation generation approach based on population-based optimization algorithms. It is mainly divided into two parts. First, it puts forward a role modelling approach based on dynamic selfadaptive genetic algorithm and NURBS technology. Second, following the introduction to PSO (Particle Swarm Optimization) algorithm, a group path generative approach is presented. It simulates group behaviours, including cohesion and separation, dynamic object tracking and collision avoidance. Finally, a group of shark modelling and path generation images are exhibited as examples.