基于连续样条曲线的替身运动平滑过渡算法

网络虚拟场景中,替身作为用户在场景中的表现方式,可以在场景中运动并与其他用户的替身或事物进行交互,替身行为的逼真性直接影响到虚拟场景的逼真性。采用关节化的人体模型可以较有效地提高逼真度,但关节化的替身与非关节化的其他场景对象相比,在运动和交互过程中大大地增加了网络负载,因此需要采用DR算法来有效地减小网络通信负载。本文研究了适合于关节化人体模型的DR预测算法,分析了运用DR预测算法不可避免地产生的替身运动的不连贯性,在此基础上提出了一种基于连续样条曲线实现的平滑过渡算法,较好地解决了替身运动的平滑过渡问题。

Avatar Motion Smoothing Algorithm Based on Splined Curve

Avatar, as represence of user, could move and interact with other avatars in networked virtual environments. Whether the behavior of avatar is realistic or not will affect the reality of the whole virtual scene. So we use articulated structure of the human body to represent the participants to improve the reality degree. But the articulated structure of the human body introduces a new complexity in the usage of the network resources; this might create a significant overhead in communication. So we use DR algorithm to reduce the network overhead. In this paper, we first analyze some different DR approach for articulated virtual human figures, then according to the limitation of DR algorithm, we present a smoothing algorithm of avatar motion based on splined curve.