基于运动捕获数据的人体运动融合

为了提高运动捕获数据的重用性和重用效率,引入注册曲线思想。以多个已有的捕获数据为输入,通过运动融合生成一段新的运动数据。首先建立时间对齐曲线,基于时间对齐曲线建立坐标对齐曲线,然后基于对齐曲线和坐标对齐曲线进行姿势融合。最后进行约束处理,扩展了可以用于运动融合的运动集合。实验结果证明方法可行且效果好。

Motion Blending Based on Motion Capture Data

To improve the reuse and efficiency of motion capture data, we introduced registration curve into motion blending. took a number of existing motion capture data as input data, then created a new series of motion data by motion blending. We built a timewarp curve ,and construct an alignment curve based on timewarp curve, then combined motion clips follow the timewarp curve and alignment curve, finally, matched constraint states for input motions, it expanded the range of motions that can be successfully blended. Experimental results show that this method is feasible and effective.