基于微型传感器驱动的三维实时运动人体模型  

Micro-sensor driven human model for 3D real-time movement

作　　者：李刚[1] 吴政[1] 孟小利[1] 吴健康[1] 

机构地区：[1]中国科学院研究生院信息科学与工程学院,北京100049

出　　处：《计算机应用》2009年第12期3336-3339,共4页journal of Computer Applications

摘　　要：建立一种运用于人体运动捕获与再现系统的人体模型,实现人体运动实时控制模型的运动。根据微型传感器人体运动捕获与再现系统的特点,提出一种三维人体模型的层次化结构建模方法和多边形分组皮肤变形方法。该方法以生物动力学为基础,根据人体关节的运动特征进行分层表示,采用传感器数据驱动并控制人体骨架模型的运动以及皮肤模型的变形,实时再现真实人体运动。实验表明,新建模方法适用于传感器数据驱动人体模型的实时运动再现,并能够逼真地体现人体运动的生物力学特征。A human model for the human motion capture and reconstruction was set up and the real-time movement of human model controlled by real people＇s movement was realized. According to the features of the system of the human motion capture and reconstruction, the authors proposed a method of constructing a sensor data-driven layered human motion model and a method of polygon grouping to achieve skin deformation. The methods are based on biomechanics. The joints of human body were categorized according to the motion characteristics. Motion parameters estimated from sensor data were used to control the movement of human skeleton model and the deformation of skin model, and consequently reconstruct the human body motion in real-time animation. The experimental results show that the proposed model is suitable for sensor data-driven virtual-human in real-time animation and has good biomechanics characteristics in representing human motion.

关 键 词：人体模型 四元数 微型传感器 实时动画 运动再现 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]