首页 | 本学科首页   官方微博 | 高级检索  
     

基于内距离形状上下文的跌倒检测方法
引用本文:王亚飞,杨庚,李百惠. 基于内距离形状上下文的跌倒检测方法[J]. 计算机技术与发展, 2014, 0(3): 58-62
作者姓名:王亚飞  杨庚  李百惠
作者单位:南京邮电大学 计算机学院,江苏 南京210003
基金项目:国家自然科学基金资助项目(61272084,61202004);江苏省高校自然科学研究重大项目(11KJA520002);江苏省科技支撑计划(社会发展)项目(BE2011826);高等学校博士学科点专项科研基金资助课题(20113223110003,20093223120001)
摘    要:
在全球社会老龄化的大背景下,老年人的身体健康状况和晚年生活质量需要更多的关注。跌倒在老年人群中发生率高并且带来的后果比较严重。文中提出一种应用于家庭场景的基于Inner-Distance形状上下文( Inner-Distance Shape Context,IDSC)的跌倒检测方法。该方法通过Inner-Distance形状上下文获得视频帧前景形状的描述信息,使用形状匹配方法对视频序列中人体形状变化进行量化。对形变量化信息使用动态时间规整( Dynamic Time Warping,DTW)方法实现跌倒行为的判定。实验结果表明该方法可有效、快速地判断跌倒,提取结果具有较好的查准率和查全率。

关 键 词:跌倒检测  形状上下文  动态时间规整

Fall Detection Approach Based on Inner-distance Shape Context
WANG Ya-fei,YANG Geng,LI Bai-hui. Fall Detection Approach Based on Inner-distance Shape Context[J]. Computer Technology and Development, 2014, 0(3): 58-62
Authors:WANG Ya-fei  YANG Geng  LI Bai-hui
Affiliation:( College of Computer Science & Technology, Nanjing University of Posts &Telecommunications, Nanjing 210003, China)
Abstract:
Faced with the growing population of seniors,the society needs to pay more attention to the life quality and health condition for seniors. What's more,falls in the elderly population have a high incidence and a more serious consequence. An adaptive fall detection ap-proach based on inner distance shape context in home environment is presented. This method is based on analyzing human shape deforma-tion during a video sequence. A inner distance shape context method is used to track the person's silhouette along the video sequence. The shape deformation is then quantified from these silhouettes based on shape analysis methods. Finally,falls are detected from normal activities using dynamic time warping methods. Experiments show that the fall detection approach proposed can detect the falls effectively and rapidly,the results have good precision and recall ratio.
Keywords:fall detection  inner distance shape context  dynamic time warping
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号