...
首页> 外文期刊>Pattern Analysis and Applications >Touchless typing with head movements captured in thermal spectrum
【24h】

Touchless typing with head movements captured in thermal spectrum

机译:非接触式打字,可在热光谱中捕获头部动作

获取原文
获取原文并翻译 | 示例
   

获取外文期刊封面封底 >>

       

摘要

Many physically challenged people are unable to operate standard electronic equipment or computer input devices. They need special assistive technologies and one of the options is the head operated interface. Face-oriented algorithms often assume a particular level of lighting with adequate intensity and spatial configuration. In the paper, we propose a thermal-imaging-based algorithm of head operating typing. It does not assume the visible light illumination. We investigated, in context of thermal imagery, several contemporary general purpose object detectors known to be accurate in case of images captured by the visible light camera. Then, a selected face detector is employed in the head operated interface analysing head movements in the thermal spectrum. The attention has been focused on the problem of touchless typing which is performed in the existing solutions either through the camera mouse or through traverse procedure with an addition mechanism (like eye blink or mouth open) needed for clicking events in both cases. Our novel solution for touchless typing with head movements combines the thermal imaging for capturing user action with the hierarchical letter selection procedure. The solution employed allows to reach any alphabet character in just three steps, i.e. with directional head movements, without the need of any additional mechanisms for clicking events.
机译:许多残障人士无法操作标准的电子设备或计算机输入设备。他们需要特殊的辅助技术,并且其中之一是头部操作界面。面向面部的算法通常假设具有足够强度和空间配置的特定级别的照明。在本文中,我们提出了一种基于热成像的头部操作打字算法。它不假定可见光照明。我们在热成像的背景下研究了几种现代通用目标探测器,这些探测器在已知由可见光相机捕获的图像中是准确的。然后,在头部操作界面中使用选定的面部检测器来分析头部在热频谱中的运动。注意力集中在非接触式打字的问题上,这是在现有解决方案中通过相机鼠标或通过遍历过程以及两种情况下都需要单击事件的附加机制(例如眨眼或张开)来执行的。我们针对头部移动的非接触式打字的新颖解决方案结合了用于捕获用户动作的热成像和分层字母选择程序。所采用的解决方案允许只用三个步骤就可以到达任何字母字符,即具有方向性的头部移动,而无需任何其他的用于单击事件的机制。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号