人体运动数据重构方法进展*

运动捕捉系统产生的人体运动数据是标记点在运动序列中的位置数据，用于驱动人体模型产生真实感的动画。在对近几年有关人体运动数据重构的文献进行综合和分析的基础上，首先对人体运动数据重构进行了问题描述，并对人体运动数据在重构过程中难以避免的噪声问题和特征点(虚拟空间中的标记点)缺失问题的研究分别进行了总结和分析；然后对人体运动数据获取的光学式原型捕捉系统开发的研究进行了讨论，评述了人体运动数据驱动人体几何模型的相关研究；最后对未来研究提出了一些展望。

Survey of human motion data reconstruction methods

Human motion data created from motion capture system are marker position data of motion sequence, and these data drive virtual human model to form an authentic animation. Based on the recent literatures of human motion data reconstruction, this paper attempted to make a survey of human motion data reconstruction methods. First,described human motion data reconstruction. Then summarizd and analyzed the research works about the unavoidable noise and feature point missing problem during the reconstruction of human motion data respectively. Subsequently,discussed the explorations of original optical capture system for capturing human motion data. Then reviewed the research findings about driving human geometrical model with human motion data. At the end, brought forward the future works for the next study.