A Real-Time Human Imitation System Using Kinect
A Real-Time Human Imitation System Using Kinect
复制标题
使用 Kinect 的实时仿人系统
DOI：
10.1007/s12369-015-0296-9
复制
发表时间：
2015-11-01
影响因子：
4.7
通讯作者：
Li, Xiaoyun
中科院分区：
文献类型：
--
作者：
Ou, Yongsheng;Hu, Jianbing;Li, Xiaoyun
机器人通常用于人类难以执行危险任务的恶劣环境中。机器人模仿为类人机器人提供了一种真实的模仿人类行为的自然方法。然而，它从来没有一个简单的任务，以建立一个有效的和强大的模仿系统，因为涉及运动的高度自由度。在本文中，我们提出了一个仿人系统的基础上Aldebaran NAO机器人和微软Kinect，它可以模仿整个身体的动作在真实的时间。通过优化过程求解逆运动学，运动被分解成由机器人关节角度列表表示的关键帧。然后通过NAO中关节电机的控制导出每个关节角度。此外，还考虑了单支撑和双支撑阶段的平衡保持以及自碰撞的避免。实验结果表明，该系统具有较强的鲁棒性和灵活性，能够很好地模拟人体的各种运动。
Robots are commonly used in harsh environments where it is difficult for humans to carry out dangerous tasks. Robot imitations provide a natural approach for humanoid robots to mimic the behavior of humans in real time. However, it is never an easy task to build an efficient and robust imitation system because of the high degree of freedom involved in motions. In this paper, we propose a human imitation system based on the Aldebaran NAO robot and the Microsoft Kinect, which can mimic the motions of the whole body in real time. By solving inverse kinematics through an optimization process, motions are split up into critical frames which are represented by a list of robot joint angles. Each joint angle is then derived through the control of the joint motors in NAO. In addition, balance maintenance in both the single and double supporting phases as well as the self-collision avoidance are taken into consideration. Experimental results show that the system is robust and flexible enough to imitate various human motions.