Li Wenda, Tan Bo, Piechocki Robert
CSN laboratory, School of Computer Science, Electrical and Electronic Engineering, and Engineering MathematicsUniversity of BristolBristolBS8 1UBU.K.
School of Computing, Electronics, and MathsCoventry UniversityCoventryCV1 5FBU.K.
IEEE J Transl Eng Health Med. 2018 Jan 25;6:2800210. doi: 10.1109/JTEHM.2018.2791609. eCollection 2018.
This paper proposes a passive Doppler radar as a non-contact sensing method to capture human body movements, recognize respiration, and physical activities in e-Health applications. The system uses existing in-home wireless signal as the source to interpret human activity. This paper shows that passive radar is a novel solution for multiple healthcare applications which complements traditional smart home sensor systems. An innovative two-stage signal processing framework is outlined to enable the multi-purpose monitoring function. The first stage is to obtain premier Doppler information by using the high speed passive radar signal processing. The second stage is the functional signal processing including micro Doppler extraction for breathing detection and support vector machine classifier for physical activity recognition. The experimental results show that the proposed system provides adequate performance for both purposes, and prove that non-contact passive Doppler radar is a complementary technology to meet the challenges of future healthcare applications.
本文提出将无源多普勒雷达作为一种非接触式传感方法,用于在电子健康应用中捕捉人体运动、识别呼吸和身体活动。该系统利用现有的家庭无线信号作为源来解读人类活动。本文表明,无源雷达是一种适用于多种医疗保健应用的新颖解决方案,可对传统智能家居传感器系统起到补充作用。文中概述了一种创新的两阶段信号处理框架,以实现多功能监测功能。第一阶段是通过高速无源雷达信号处理来获取初步的多普勒信息。第二阶段是功能信号处理,包括用于呼吸检测的微多普勒提取和用于身体活动识别的支持向量机分类器。实验结果表明,所提出的系统在这两个方面均具有足够的性能,并证明非接触式无源多普勒雷达是一种可应对未来医疗保健应用挑战的补充技术。