计算机工程与应用 ›› 2022, Vol. 58 ›› Issue (5): 50-65.DOI: 10.3778/j.issn.1002-8331.2109-0393

• 热点与综述 • 上一篇    下一篇

老年人跌倒检测算法的研究现状

赵珍珍,董彦如,曹慧,曹斌   

  1. 1.山东中医药大学 智能与信息工程学院,济南 250355 
    2.山东中医药大学附属医院 心病科,济南 250000
  • 出版日期:2022-03-01 发布日期:2022-03-01

Research Status of Elderly Fall Detection Algorithms

ZHAO Zhenzhen, DONG Yanru, CAO Hui, CAO Bin   

  1. 1.School of Intelligence and Information Engineering, Shandong University of Traditional Chinese Medicine, Jinan 250355, China
    2.Department of Cardiology, Affiliated Hospital of Shandong University of Traditional Chinese Medicine, Jinan 250000, China
  • Online:2022-03-01 Published:2022-03-01

摘要: 近些年,老年人的健康问题越来越受到重视,跌倒作为影响老年人健康安全问题的主要原因之一,其研究热度一直居高不下,高质量的跌倒检测算法层出不穷。总结了跌倒检测的研究意义和现有的热门研究方法,分别从单一算法和混合算法的角度概述基于阈值、机器学习与深度学习三个方面的跌倒检测算法,介绍各算法的检测方式、判定方式、总体性能和各类单一算法的优缺点,并且从时间、空间和时空三重维度重点阐述了卷积神经网络在跌倒领域发挥的显著作用及应用;同时介绍了跌倒检测算法所使用的数据集及其特点,便于研究者了解跌倒检测在阈值、机器学习与深度学习方面的最新研究进展。最后,对跌倒检测算法所面临的挑战及未来发展进行了展望。

关键词: 跌倒检测, 卷积神经网络, 阈值, 机器学习

Abstract: In recent years, the health problems of the elderly have been paid more and more attention. Fall is one of the main reasons that affect the health and safety of the elderly. The research interest has always been high, and high-quality fall detection algorithms have emerged one after another. This article summarizes the research significance of fall detection and existing popular research methods, and then summarizes fall detection algorithms based on a threshold, machine learning, and deep learning from the perspective of a single algorithm and hybrid algorithm. Then it introduces the detection methods of each algorithm, judgment methods, overall performance, and the advantages and disadvantages of various single algorithms, and focuses on the prominent role and application of convolutional neural networks in the field of falls from the three dimensions of time, space, and spacetime. At the same time, the data set used by the fall detection algorithm and its characteristics are introduced, so that researchers can understand the latest research progress of fall detection in terms of threshold, machine learning, and deep learning. Finally, the challenges and future development of the fall detection algorithm have prospected.

Key words: fall detection, convolutional neural networks;threshold, machine learning