欢迎访问《图学学报》 分享到:

图学学报

• 交互设计与虚拟现实 • 上一篇    下一篇

基于MEMS 传感器和Unity3D 的人体运动捕获系统

  

  • 出版日期:2015-04-30 发布日期:2015-06-03

Human Motion Capture System Based on MEMS Sensors and Unity3D

  • Online:2015-04-30 Published:2015-06-03

摘要: 随着人机交互技术的发展,人与计算机之间自然、多模态交互将成为人与计算机
之间交互的主要方式,而这首先需要计算机可以正确地理解和捕捉人的行为特征,运动捕获技
术正是在这种背景下提出来。通过运动捕获技术,计算机可以理解人体动作,用户就可以通过
体态、方位、手势和表情等模态向计算机发出指令、传达信息等,因此运动捕获是新一代人机
交互中的关键技术之一。目前基于MEMS 惯性传感器的动作捕获研究主要针对动漫和电影制
作,价格昂贵;随着传感器的集成度变高,价格逐渐降低带来了新的系统设计需求。在原有动
作捕获技术基础上,设计并实现一种普适性更好的人体动作捕获系统原型,原型系统实现了基
于惯性传感节点的人体运动信息的采集与融合、节点与汇聚节点的数据传输、虚拟人体模型实
时动作呈现程序;实现了从数据测量到采集,再到模拟呈现的全过程。

关键词: MEMS 惯性传感器, 人机交互, 动作捕获

Abstract: A natural, flexible way for human beings to interact with computers is main goal to
human-computer interaction technology, which is remarkably developing. This goal requires
computers capture and understand humanʹs motion precisely, under this background, motion capture
technology was proposed. As future goal, computer users could send information, commands to
computers not only by clicking and typing, but also by gesture, facial expression and other body
expressions. To achieve this, motion capture technology is definitely indispensable. Currently existing
motion capture systems mostly work for 3D video, which have high costs. However, the improving
integration of MEMS sensor brings lower cost and more demands to motion capture systems. With
integrating existing technologies, this paper proposes a universal motion capture prototype, which
implements the complete process from data measuring, transmitting to motion simulating presence
based on inertial sensor nodes.

Key words: MEMS sensor, human-computer interaction, motion capture