摘要
设计了一款基于Android便携式智能老年健康监护系统,系统包括跌倒检测、心率检测、地图定位3大功能模块。跌倒检测模块使用最大类间方差法计算最优阈值,采用合加速度幅值面积的曲线相似度区分跌倒行为与较高强度日常活动;采用小波阈值去噪算法去除输入信号中的噪声项,获得准确的实时心率测量结果;采用地图API的定位SDK,及时提供精确的位置服务,将异常警告及位置信息实时发送至监护人手机。系统采用Java语言进行设计开发,在Andorid平台上运行稳定。测试结果表明:跌倒检测的误报率和漏报率分别为1. 67%、2%,心率检测的平均绝对误差和平均相对误差分别为1. 6 bpm、2. 08%,可满足老年人监护的需求。
A portable intelligent health monitoring system based on Android is designed. This system includes three functional modules: fall detection,heart rate detection and map location. In the fall detection module,the maximum of inter-class variance method is used to calculate the optimal threshold,and the curve similarity of the area of the acceleration amplitude is used to distinguish fall behaviors from high intensity daily activities. Thewavelet threshold de-noising algorithm is used to eliminate noise items in input signals,and accurate real-time heart rate measurement results can be obtained. The SDK location based on map API can provide precise location service in time and send abnormal warning and location information to a guardian ’s cell phone in real time. The system designed and developed in Java language runs stably on Android platform. The test results show that the false alarm rate is 1. 67% and missing alarm rate is 2% regarding fall detection while the average absolute error is 1. 6 BPM and the relative error is 2. 08% regarding heart rate detection,which can satisfy the needs of elderly people’s guardianship.
作者
岳佳欣
王忠
郑晓彬
YUE Jiaxin;WANG Zhong;ZHENG Xiaobin(Sichuan Film and Television University,Chengdu 610036,China;College of Electrical Engineering and Information,Sichuan University,Chengdu 610065,China)
出处
《重庆理工大学学报(自然科学)》
CAS
北大核心
2020年第4期200-208,共9页
Journal of Chongqing University of Technology:Natural Science
基金
四川省高等学校人文社会科学重点研究基地-新建院校改革与发展研究中心项目(XJYX2019B01)
四川省教育厅2018自然科学重点科研项目(18ZA0307、18ZA0308)
四川省科技厅科技支撑项目(2015FZ061)。
关键词
老年人智能监护
ANDROID手机
跌倒检测
心率检测
intelligent monitoring for senior citizens
android phone
fall detection
heart rate detection