基于视觉识别和多传感器的跌倒检测系统设计  

Fall detection system design based on visual recognition and multi-sensor

作　　者：程世通 张李辉 楚遵恒 余韵 李小雨 刘紫燕 CHENG Shitong;ZHANG Lihui;CHU Zunheng;YU Yun;LI Xiaoyu;LIU Ziyan(College of Big Data and Information Engineering,Guizhou University,Guiyang 550025,China;State Key Laboratory of Public Big Data,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学大数据与信息工程学院,贵州贵阳550025 [2]贵州大学公共大数据国家重点实验室,贵州贵阳550025

出　　处：《传感器与微系统》2024年第8期91-94,共4页Transducer and Microsystem Technologies

基　　金：贵州省2020年大学生创新创业训练计划项目(S202010657061);贵州省联合资金资助项目(黔科合LH字[2017]7226号);贵州大学2017年度学术新苗培养及创新探索专项项目(黔科合平台人才[2017]5788)。

摘　　要：针对老人意外跌倒摔伤无法及时得到救助的问题,设计了一个基于视觉识别和多传感器的跌倒检测系统。系统以树莓派为核心,采用摄像头、传感器、定位模块、心率模块和窄带物联网(NB-IoT)模块采集和传输信息,实现了老人室外和室内跌倒检测。利用阈值检测和支持向量机(SVM)分类的方法提高了跌倒检测的准确率。利用监护App实现了老人状态和位置信息查询功能。实验结果表明:该系统室外和室内跌倒检测正确率分别为92.8%和91.0%。该系统能有效检测老人的跌倒行为并给监护人发送报警信息和定位信息,实现了智能看护老人。Aiming at the problem that accidentally fall of elderly can not be rescued in time,a fall detection system based on visual recognition and multi-sensor is designed.Taking Raspberry Pi as the core,the proposed system,cameras and sensors,positioning module,heart rate module,and narrow band Internet of Things(NB-IoT)module are used to collect and transmit information and the elderly fall detection indoor and outdoor are accomplished.Threshold detection and SVM classification method are used to improve fall detection accuracy.Monitoring App is used to realize the elderly status and location information inquiry function.The experimental results demonstrate that the accuracy of fall detection indoors and outdoors is 92.8%and 91.0%,respectively.The system can effectively detect the elderly fall behavior and send alarm and positioning information to their supervisors,which provide wise care for the elderly.

关 键 词：跌倒检测 支持向量机 阈值检测 OpenPose算法 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]