摘要
提出一种新的人机交互方法并设计了系统原型,通过单个摄像头跟踪人手在空中的运动轨迹,生成相应的字符图像,再结合特征提取和分类器算法识别所写字符。实验结果表明,该系统能够实现实时手部跟踪和手写识别的功能,其设计思想可用于虚拟现实、沉浸式游戏等应用领域。
This paper proposes a new method for human-machine interaction and designs a prototype system for it.By tracking the movement of human hand,it generate the path image comprising character information corresponding to the track of hand,and combine with feature extraction algorithm and classification system it can recognize the handwritten character.Experimental results prove the validity of this method,and the idea can be applied to virtual reality.
出处
《软件导刊》
2012年第7期87-89,共3页
Software Guide
基金
武汉科技大学大学生创新基金项目(10ZRP060)
关键词
人机交互
单目视觉
目标跟踪
手写识别
Human-Machine Interaction
Monocular Vision
Object Tracking
Handwriting Recognition