一种光学式运动捕捉系统下的物体运动数据捕捉方法研究

Research on a method of capturing rigid body motion characteristicsof optical motion capturing system

摘要:

多相机视觉运动捕捉系统能通过捕捉标记点的空间坐标从而获得运动物体的运动学参数，本文提出了一种基于多相机运动捕捉系统下的通用物体运动数据捕捉方法。首先根据三个标记点组成固定模型获取物体运动过程中对应标记点的瞬时坐标，然后通过向量法求解出被测物体在运动过程中各采集点对应的物体位姿，然后通过卡尔曼滤波方法消除运动捕捉过程中的系统和环境误差的影响，获得平滑的物体位姿运动轨迹，并根据滤波数据计算出物体在各采集点对应的速度、加速度、角速度、角加速度。最后基于协作机器人进行物体的运动数据捕捉实验，验证了所提出物体运动数据捕捉方法的正确性。

Abstract:

The multi-camera visual motion capture system can obtain the kinematic parameters of the moving object by capturing the space coordinates of the marked points. This paper proposes a general object motion data capture method based on the multi-camera motion capture system. First, obtain the instantaneous coordinates of the corresponding mark points during the movement of the object according to the fixed model composed of three mark points, and then use the vector method to solve the object pose corresponding to each collection point of the measured object during the movement, and then eliminate it by the Kalman filter method. Due to the influence of system and environmental errors in the motion capture process, a smooth object pose movement trajectory is obtained, and the corresponding velocity, acceleration, angular velocity, and angular acceleration of the object at each collection point are calculated according to the filtered data. Finally, the object motion data capture experiment based on the collaborative robot verifies the correctness of the proposed object motion data capture method.