基于智能体的群体动画创作方法的研究与实现

当前，动画研究日益着重角色的智能和动作的细腻，群体动画系统的各模块之间耦合度日益提高，致使动画创作难度日益加大，尤其当智能角色数目增多时，其计算量将呈非线性快速增加，极大地影响了动画创作的实时性，无法在简单性和高效性方面较好地均衡。该文基于Steering行为理论，综合传统的智能角色动画和蒙皮骨骼动画技术，提出了一个基于自主智能体的群体动画角色模型。实例证明：此种模型制作简单，模块耦合度低且能通过很少的计算量实现实时交互，提高了动画创作效率，具有良好的应用前景。

Research and Implementation of Group Animation Based on Autonomous Agents

Current animation research focuses on either artificial intelligence aspects or advanced development of the animation.The excessive coupling between modules of group animation system makes it difficult to produce animation efficiently. Especially, when having a number of autonomous roles, the total calculation will increase non-linearly too quickly to influence the real-time interaction. They all do not emphasize simplicity and efficiency in both. A group animation model based on multiple autonomous agents is presented in this paper, which is under the direction of Steering behavior theory, combining traditional autonomous role animation and the sin-skeleton animation technique. The animation instance proves this model can realize real-time interaction with light computation load, low module coupling , improves animation production efficiency and will have a wide application in the areas such as entertainment, virtual reality and animation.