文献知网节
  • 记笔记

Algorithm Implementation of Android-based Fall-detecting System

Meihui LiGaosheng XieGuoru ZhaoXing GaoZongzhen JinYingnan Ma

Shenzhen Institutes of Advanced Technology,Chinese Academy of SciencesWuhan Research Institute of Post and TelecommunicationsBeijing Research Center of Urban System Engineering

摘要:As one of the main causes of accidental injury death, falls not only brings a lot of health problems, but also caused huge economic losses. And accidental falls are the main factors that endanger the health of elderly people in modern society. Timely and effective fall detection and alarm can reduce the risk of falls. Through the collection, research and analysis of the data of human’ behaviour, study the fall detection and prevention program to promote the standardization and scientific of prevention and treatment of falls. It has far-reaching significance on protecting human’s health and improving the overall quality of medical care. Along with the increasing of the aging society, the design and development of the fall detection system which is portable, accurate and real-time has gradually become one of the most urgent needs of the community. At present the fall detection system has realized the functions of real-time detection in a specific environment, but it still cannot meet the actual needs because of the disadvantages such as high cost, limited monitoring scope, easy to be influenced by environment, and invasion of privacy, inaccurate and inefficient, etc.. With the rapid development of mobile Internet, smart phones as another kind of wearable device have become essential products of people’s life. Smart phone is portable and flexible, and built in various sensors which can be used to monitor human motion data. In order to promptly fall detect and minimize the damage to the aged, a fall detection system based on Android phones is designed and implemented and a fall detection algorithm for phones is proposed in the paper. Firstly, it collects real-time sensor data through the phone’s built-in sensors, analyses and processes the data and extracts data features. And then it analyses these human motion parameters and movement characteristics, researches and builds a fall detection model. By comparing and analysing the advantages and disadvantages of the linear Supported Vector Machine(SVM) and the Decision Tree algorithm, it makes a concrete analysis of the question and proposes the two detection algorithm based on SVM and Decision Tree algorithm to improve the accuracy and sensitivity of detection. There are two steps to detect fall with this algorithm. The first one is used to detect the daily behaviour by one dimensional model of SVM. If the result is falling, and then the second Decision Tree detection to distinguish falls finally. While falls, the mobile telephone automatically alarms so that the elderly people can get rescue in time. Comparing with stand-alone SVM and Decision Tree algorithm, it simplified the dimension of the model. And it is better adapted to the mobile phone carrier, simplifies the detection process of daily, and effectively improves accuracy and sensitivity of the system by two fall detection processes. In this way, the elderly can detect falls without other equipment at any time. Meanwhile, the system integrates t
会议名称:

第五届国际仿生工程学术大会(ICBE2016)

会议时间:

2016-06-21

会议地点:

中国浙江宁波

  • 专辑:

    信息科技

  • 专题:

    自动化技术; 自动化技术

  • 分类号:

    TP181;TP212.9

  • 手机阅读
    即刻使用手机阅读
    第一步

    扫描二维码下载

    "移动知网-全球学术快报"客户端

    第二步

    打开“全球学术快报”

    点击首页左上角的扫描图标

    第三步

    扫描二维码

    手机同步阅读本篇文献

  • CAJ下载
  • PDF下载

下载手机APP用APP扫此码同步阅读该篇文章

下载:2 页码:156 页数:1 大小:205k

引文网络
  • 参考文献
  • 引证文献
  • 共引文献
  • 同被引文献
  • 二级参考文献
  • 二级引证文献
  • 批量下载
相关推荐
  • 相似文献
  • 读者推荐
  • 相关基金文献
  • 相关法规
  • 关联作者
  • 相关视频