摘要
姿势识别在三维虚拟实验中有着重要的作用,其作用体现在通过特定的姿势,控制程序准确地响应某一个操作来完成所对应的实验中的特定功能。为了便于使用,同时使得姿势识别的准确率提高,文中提出了利用Kinect传感器得到的二十多个关节点信息,通过确定每个关节点之间的欧氏距离与角度来判别特定姿势的方法。结果表明,使用此方法来识别姿势,识别率较高,且可随时扩展动作库来满足判定不同姿势的需求。本算法计算较为简单,效率高,能满足日常的虚拟实验要求。其便利的姿势库扩展性,有利于软件复用,提高虚拟实验的使用效率。
Gesture recognition plays an important role in the three-dimensional virtual experiments, its role is to control program in re- sponding to an action by a particular posture. For ease of use and making the gesture recognition more accurate, propose the use of Kinect sensor to get over twenty articulation points information, and determine the Euclidean distance and angle of the articular point to distin- guish a particular posture. The results show that using this method can always expand action library to meet the needs of different postures of the decision. The algorithm is relatively simple to calculate and high efficiency,they are able to meet the day-to-day requirements of the experiment. It＇ s helpful to software reuse,and improves the efficiency in the use of virtual experiment.
出处
《计算机技术与发展》
2013年第5期258-260,共3页
Computer Technology and Development
基金
广州市科信科技攻关项目(20121208001A)
广州市2012年高校科研计划项目(2012A095)
