基于数字图像处理的虚拟输入方法  

Virtual input method based on digital image processing

作　　者：张金[1] 李洋[1] 刘晓威[1] 庞赫洋[1] 侯洋洋[1] 

机构地区：[1]南开大学信息技术科学学院,天津300071

出　　处：《计算机工程与设计》2011年第11期3759-3763,3919,共6页Computer Engineering and Design

基　　金：国家大学生创新性实验计划基金项目(101005514)

摘　　要：针对当前嵌入式设备输入方法存在的高成本低体验问题,提出了一种运行于嵌入式系统中基于数字图像处理的虚拟输入方法。该方法利用嵌入式设备附加的摄像设备,对用户在虚拟键盘上的按键过程进行图像采集,通过角点识别和图像分割进行按键识别,以便通过采集到的图像定位出用户点击的目标按键,从而实现信息输入过程。通过实例验证了该方法的可行性和有效性,为嵌入式设备的交互设计提供了可供参考的处理方法。Aimed at the problems oflow cost-effective in embedded input systems, the virtual input method in embedded system based on digital image processing is presented. In this method, user typing process is recorded as motion jointed images by the camera in embedded device. After several image processing steps, such as comer reorganization and image segmentation, the target key clicked is located. Repeating the process, the information is transformed into device. Finally, the feasibility of the method is validated by practical application. It provides a reference approach for human-computer interaction design in embedded fields.

关 键 词：嵌入式系统 虚拟输入方法 伪键盘 嵌入式设备 人机交互 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]