摘要
为应对医院、敬老院及一些特定场所的独居老人跌倒问题,做好特殊人群跌倒的检测至关重要。采用新兴的基于Wi-Fi的信道状态信息,用FallBR(fall behavior recognition)方法,专门针对跌倒与跌倒相似行为(下蹲、弯腰拾物)进行识别。FallBR采用基于小波特征与时频特征相结合的方法,滤去人员行走带来的干扰,采用基于PCA的载波选择方法实现无需训练的跌倒行为的有效识别,相较其它方法,提高了跌倒识别的准确度,降低了假阳率。
To cope with the fall risk of old people living alone in the hospital,the elderly hospital and some specific places,it is very important to detect the fall of special people.The new channel state information based on Wi-Fi was adopted and the FallBR (fall behavior recognition) method was used to detect the similar behaviors of fall (squat,bend down and pick up things ).FallBR adopted a method based on the combination of wavelet feature and time frequency feature to filter out the interference caused by people walking,and the carrier selection method based on PCA was used to realize the effective recognition of the fall without training.Compared with other methods,the accuracy of the fall recognition is improved and the false positive rate is reduced.
作者
崔然
冯秀芳
CUI Ran;FENG Xiu-fang(College of Information and Computer,Taiyuan University of Technology,Taiyuan 030024,China)
出处
《计算机工程与设计》
北大核心
2019年第8期2331-2336,共6页
Computer Engineering and Design
基金
国家自然科学基金项目(61472272)
关键词
信道状态信息
菲涅尔区模型
小波变换
时频分析
主成分分析
channel state information
Frenel model
wavelet transform
time-frequency analysis
principal component analysis
作者简介
崔然(1994-),男,山西运城人,硕士研究生,研究方向为无线传感网、物联网;冯秀芳(1966-),女,山西太原人,教授,硕士生导师,CCF高级会员,研究方向为无线传感网、人工智能等。E-mail:cuiran5215@qq.com.
