• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Sensor Letters: A Journal Dedicated to all Aspects of Sensors in Science, Engineering, and Medicine >A Novel Video Sensor Based Fall Detection of the Elderly Using Double-Difference Image and Temporal Templates
【24h】

A Novel Video Sensor Based Fall Detection of the Elderly Using Double-Difference Image and Temporal Templates

机译：基于双图像和时间模板的基于视频传感器的老年人跌倒检测

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In recent years, fall incident detection is one of the major health care issues because unexpected falls cause serious injuries in elderly people. Several related studies based on video sensor have tried to detect falls, however, their methods offer low fall detection rates in general. To overcome this problem, we propose a novel approach to detect falls using a weighted subtraction between consecutive difference images and a motion history image on temporal templates in real time. As a result, the proposed algorithm obtains the successful rate of 96% even though the video sequence is obtained by an USB PC camera sensor. In addition, the sensitivity and specificity of our system are 95% and 97.2%, respectively. Experimental result also shows advanced reliability to discriminate rapid sitting down from an abrupt fall which is most difficult to detect. Therefore, Our PC camera sensor-based algorithm could be used for fall detection and activity monitoring in elderly people with high reliability and resolution.
机译：近年来，跌倒事件检测是主要的医疗保健问题之一，因为意外跌倒会对老年人造成严重伤害。一些基于视频传感器的相关研究试图检测跌倒，但是，他们的方法通常提供较低的跌倒检测率。为了克服这个问题，我们提出了一种新颖的方法来检测跌倒，方法是使用连续差分图像和时间模板上的运动历史图像之间的加权减法实时进行。结果，即使视频序列是通过USB PC摄像机传感器获得的，所提出的算法也获得了96％的成功率。此外，我们系统的敏感性和特异性分别为95％和97.2％。实验结果还显示出先进的可靠性，可将快速坐下与突然跌倒区分开，这是最难发现的。因此，我们的基于PC相机传感器的算法可用于具有高可靠性和分辨率的老年人跌倒检测和活动监测。

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