Leap Motion关键点模型手姿态估计方法

Key Point Model for Hand Pose Estimation Based on Leap Motion

在虚拟操作的人机交互中,为解决视觉方式人手姿态估计的问题,提出一种基于Leap Motion采集设备的关键点模型手姿态参数估计方法.该方法通过建立关键点模型,利用Leap Motion采集的少量手势信息计算模型中人手关键点的空间位置,并将其作为手姿态的估计参数.实验结果表明,文中方法能够实时地估算出手各个关键点的位置以驱动虚拟手运动,为基于手势的虚拟操作人机交互应用奠定了基础.

In order to solve the problems of vision-based hand pose estimation in human-computer interactive of virtual operations, this paper proposed a key point model for pose estimation based on Leap Motion. With the information obtained from Leap Motion, the proposed model calculated the positions of the key points as the parameters of hand pose. Our experiments showed that this estimation method could obtain the positions of key points in real time to drive virtual hands in virtual operation interactive based on hand gestures.