养老机器人的跌倒行为检测方法研究

A Review of Researches on Falling Behavior Detection Methods Used in Pension Robots

随着人口老龄化的形势不断严峻,养老成为社会一个焦点话题。大多数年轻人选择背井离乡或者疲于工作,造成空巢老人数量不断增加。而在老人独处时发生安全隐患概率明显增大。本文以计算机视觉为基础,探讨老人独处时发生跌倒行为的现象的解决方案,介绍了特征提取和深度学习两种主流思路,详细说明了其发展和常用方法,同时对公开数据集进行利弊分析。最后总结了全文使用的方法同时阐述了对该技术未来发展方向的见解。

As the population ageing situation continues to be severe,elderly care has become a hot topic in society.Most young peo⁃ple choose to leave their homes or get tired of work,resulting in an increasing number of empty-nest elderly people.The probabili⁃ty of potential safety hazards when the elderly is alone increases significantly.Based on computer vision,the solution to the phe⁃nomenon of falling behavior when the elderly is alone is explored,two mainstream ideas of feature extraction and deep learning are introduced,and its development and common methods are explained in detail.At the same time,the pros and cons of public data sets are analyzed.Finally.Summarizes the methods used in the full text and expounds insights into the future development direc⁃tion of the technology.