老年人摔倒检测与预警系统的设计与实现
发布时间：2018-07-17 06:38
【摘要】：摔倒已经成为目前最主要的导致老年人受伤以及住院的意外事件。对老年人摔倒事件的及时发现能有效的降低老年人受到的损伤。由于智能手机的普适性,基于智能手机的摔倒检测与预警系统成为近些年来的研究热点,但是以往的研究往往只利用了加速计一个手机元件来作为摔倒检测的依据,因此如何充分利用智能手机上已有的传感器来设计一个高检测率、低误报率的系统是本文的研究重点。本文的研究目的在于利用从智能手机及智能手表提取出的设备姿态数据、相对高度数据、心率数据等设计一个高准确率、低误报率以及多预警等级的摔倒检测与预警算法,并采取多种预警方式来向看护人员发送救护预警信息,最后设计和实现了一个较为完备的老年人摔倒检测与预警系统。首先,为了利用智能手机判定老年人摔倒事件的发生,对基于智能手机的摔倒检测方法及预警方式进行研究,提出本文采取的摔倒事件检测思想,明确系统相关使用角色。在此基础上对老年人摔倒检测与预警系统进行需求分析,确定系统必备的功能需求、核心业务及系统的工作流程。其次,为了提高摔倒检测的准确率,本文基于DeviceMotion提出一种描述不同运动事件的特征向量,为了验证4种经典分类算法基于该特征向量检测摔倒事件的可靠性,本文邀请志愿者佩戴有安装运动数据采集程序的手机进行模拟摔倒试验以及每日的正常活动,基于构造的样本集对4种分类算法的性能表现进行了分析对比实验。再次,对老年人摔倒检测与预警系统进行了设计工作。遵循软件工程的思想对系统进行了设计,并在摔倒检测与预警模块中采用震动提醒、铃声提醒以及语音识别三种人机交互方式来降低摔倒事件的误报率。最后,实现并测试了老年人摔倒检测与预警系统。本文通过Core Motion框架来获取老年人的运动数据,通过Health Kit框架来获取老年人的心率数据,通过科大讯飞开源框架来实现语音识别功能,通过Core Location框架来获取老年人的位置信息,通过Cloud Kit来发送老年人的救护预警信息。最后对系统进行了功能和非功能测试。
[Abstract]:Fall has become the main cause of the injury to the elderly and the accident in hospital. The timely discovery of the fall of the elderly can effectively reduce the damage to the elderly. Because of the universality of the smart phone, the detection and early warning system based on the smartphones has become a hot research topic in recent years, but the previous research has been studied. The focus of this paper is how to make full use of the existing sensors on the smart phone to design a high detection rate and low false alarm rate. The purpose of this paper is to make use of the device posture extracted from smart phone and smart watch. The data, the relative height data, the heart rate data and so on, design a high accuracy rate, low false alarm rate and the multiple warning level fall detection and early warning algorithm, and take a variety of early warning methods to send the nursing early warning information to the caregivers, and finally design and implement a relatively complete detection and early warning system for the elderly fall. Using smart phone to determine the occurrence of falling events of the elderly, research on the detection methods and early warning methods based on the smartphone, put forward the idea of the fall event detection and clarify the related role of the system. On this basis, the needs of the elderly fall detection and early warning system are analyzed, and the necessary functions of the system are determined. Requirements, core business and the workflow of the system. Secondly, in order to improve the accuracy of the fall detection, this paper presents a feature vector describing different events based on DeviceMotion. In order to verify the reliability of the 4 classical classification algorithms based on the feature vector detection, this article invites the volunteers to wear the installed sports data. The handset of the acquisition program carries on the simulated fall test and the daily normal activities. Based on the structural sample set, the performance of the 4 classification algorithms is analyzed and contrasted. Thirdly, the design work of the elderly fall detection and early warning system is carried out. The system is designed and the fall detection is carried out following the thinking of software engineering. In the early warning module, three human computer interaction methods are used to reduce the false alarm rate of the fall events. Finally, the old people's fall detection and early warning system is realized and tested. In this paper, the Core Motion framework is used to obtain the motion data of the elderly, and the heart rate of the elderly is obtained through the Health Kit framework. The data, through the open source framework of the HKDA to realize the voice recognition function, through the Core Location framework to obtain the location information of the elderly, through the Cloud Kit to send the elderly ambulance early warning information. Finally, the system has been tested for functional and non functional.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52;TP277
[Abstract]:Fall has become the main cause of the injury to the elderly and the accident in hospital. The timely discovery of the fall of the elderly can effectively reduce the damage to the elderly. Because of the universality of the smart phone, the detection and early warning system based on the smartphones has become a hot research topic in recent years, but the previous research has been studied. The focus of this paper is how to make full use of the existing sensors on the smart phone to design a high detection rate and low false alarm rate. The purpose of this paper is to make use of the device posture extracted from smart phone and smart watch. The data, the relative height data, the heart rate data and so on, design a high accuracy rate, low false alarm rate and the multiple warning level fall detection and early warning algorithm, and take a variety of early warning methods to send the nursing early warning information to the caregivers, and finally design and implement a relatively complete detection and early warning system for the elderly fall. Using smart phone to determine the occurrence of falling events of the elderly, research on the detection methods and early warning methods based on the smartphone, put forward the idea of the fall event detection and clarify the related role of the system. On this basis, the needs of the elderly fall detection and early warning system are analyzed, and the necessary functions of the system are determined. Requirements, core business and the workflow of the system. Secondly, in order to improve the accuracy of the fall detection, this paper presents a feature vector describing different events based on DeviceMotion. In order to verify the reliability of the 4 classical classification algorithms based on the feature vector detection, this article invites the volunteers to wear the installed sports data. The handset of the acquisition program carries on the simulated fall test and the daily normal activities. Based on the structural sample set, the performance of the 4 classification algorithms is analyzed and contrasted. Thirdly, the design work of the elderly fall detection and early warning system is carried out. The system is designed and the fall detection is carried out following the thinking of software engineering. In the early warning module, three human computer interaction methods are used to reduce the false alarm rate of the fall events. Finally, the old people's fall detection and early warning system is realized and tested. In this paper, the Core Motion framework is used to obtain the motion data of the elderly, and the heart rate of the elderly is obtained through the Health Kit framework. The data, through the open source framework of the HKDA to realize the voice recognition function, through the Core Location framework to obtain the location information of the elderly, through the Cloud Kit to send the elderly ambulance early warning information. Finally, the system has been tested for functional and non functional.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52;TP277
【参考文献】
中国期刊全文数据库 前4条
1 耳玉亮;段蕾蕾;叶鹏鹏;汪媛;纪翠蓉;邓晓;高欣;金叶;王临虹;;2014年全国伤害监测系统老年跌倒/坠落病例特征分析[J];中华流行病学杂志;2016年01期
2 贾菲菲;宾光宇;吴水才;师玉涛;;基于加速度信号的纵跳力学参数计算[J];中国医疗设备;2015年09期
3 王志宝;孙铁山;张杰斐;;人口老龄化区域类型划分与区域演变分析——以中美日韩四国为例[J];地理科学;2015年07期
4 曹玉珍;蔡伟超;程e，
本文编号：2129432
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2129432.html
