基于离散特征的跌倒检测智能方法及应用
发布时间：2018-11-20 07:24
【摘要】：随着人口老龄化现象加剧,对老年人跌倒的检测与报警越来越重要。为提高跌倒检测的准确率,提出一种基于离散特征的跌倒检测智能方法。通过对人体运动数据的分析,提出7类人体运动特征;并建立了以BP神经网络为基础的跌倒检测模型,将提取的离散特征作为模型的输入,模型的输出作为跌倒检测结果;通过对模型的学习与训练后,实现跌倒检测。方法验证和产品应用结果表明:采用基于离散特征的跌倒检测智能方法能够有效地区分跌倒与非跌倒,提高了跌倒检测正确率,降低了误报率和漏报率。
[Abstract]:With the aging of the population, it is more and more important to detect and alarm the fall of the elderly. In order to improve the accuracy of fall detection, an intelligent fall detection method based on discrete features is proposed. Through the analysis of human motion data, seven kinds of human motion characteristics are proposed, and a fall detection model based on BP neural network is established. The extracted discrete features are taken as the input of the model, and the output of the model is regarded as the result of fall detection. After learning and training the model, the fall detection is realized. The results of method verification and product application show that the intelligent method based on discrete feature can effectively distinguish fall from non-fall, improve the correct rate of fall detection, and reduce the false alarm rate and false alarm rate.
【作者单位】： 后勤工程学院后勤信息与军事物流工程系;重庆市软汇科技有限公司;
【基金】：国家自然科学基金(61271449,61302175) 重庆市自然科学重点基金(CSTC2015jcyj BX0017)项目资助
【分类号】：TP183;TP274
本文编号：2344238
[Abstract]:With the aging of the population, it is more and more important to detect and alarm the fall of the elderly. In order to improve the accuracy of fall detection, an intelligent fall detection method based on discrete features is proposed. Through the analysis of human motion data, seven kinds of human motion characteristics are proposed, and a fall detection model based on BP neural network is established. The extracted discrete features are taken as the input of the model, and the output of the model is regarded as the result of fall detection. After learning and training the model, the fall detection is realized. The results of method verification and product application show that the intelligent method based on discrete feature can effectively distinguish fall from non-fall, improve the correct rate of fall detection, and reduce the false alarm rate and false alarm rate.
【作者单位】： 后勤工程学院后勤信息与军事物流工程系;重庆市软汇科技有限公司;
【基金】：国家自然科学基金(61271449,61302175) 重庆市自然科学重点基金(CSTC2015jcyj BX0017)项目资助
【分类号】：TP183;TP274
【相似文献】
相关期刊论文 前5条
1 张龙斌;贺国光;卢宇;;一种在线实时快速判定混沌的智能方法[J];天津理工大学学报;2007年01期
2 杨则正;用基于信息技术的智能方法解决复杂问题[J];管理科学文摘;1995年10期
3 王光波;曹梦龙;;基于智能方法的丝状异纤分离关键技术研究[J];工业控制计算机;2010年04期
4 池越;夏克文;陈国鹰;苏昶;;一种基于QPSO和WLS-SVM的智能方法[J];河北工业大学学报;2009年02期
5 ;[J];;年期
相关会议论文 前2条
1 张锐;赵阳;邵柏;赵一彪;马丁;;统一集智能方法在人-机-环境系统工程研究中应用[A];人-机-环境系统工程研究进展（第七卷）[C];2005年
2 付林;贺安之;;一种实时检测路面面形变的智能方法[A];江苏、山东、河南、江西、黑龙江五省光学（激光）联合学术'05年会论文集[C];2005年
，本文编号：2344238
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2344238.html
