摘要
本文在对手结构和关节运动作了深入分析的基础上，建立了对其施加约束的手模型．根据所定义的约束及运动类型，建立了适合于手势合成的手数据结构．并采用神经网络理论进行手势识别，取得了较好的效果．
This paper introduces a hand model with constraints on it based on the analysis of hand structure and joints motion. According to the defined constraints and motion type, the data structure suitable for the synthesis of hand motion has been created. And also the neural network theory has been applied to hand motion distinguishing and it works fine.
出处
《机器人》
EI
CSCD
北大核心
1999年第1期40-45,共6页
Robot
关键词
虚拟现实
手势识别
神经网络
模式识别
Virtual reality, hand model, hand motion distinguishing, neural network
