Automated Monitoring and Localization of People

拟议研究方案的长期目标是开发一个全面检测活动以及对自然环境中需要帮助的人进行生理评估的系统。例如，这包括跟踪老年人、痴呆症患者、儿童、囚犯的日常活动，以及检测老年人跌倒和心脏病发作等突发生命威胁事件。该项目将允许对独居或与其他人一起生活的人进行持续监测，或同时监测几个人。NSERC Discovery研究项目采取的方法不同于基于智能家居和对日常活动的识别的研究。本文提出了一种仅基于射频后向散射信号来跟踪和评估现实环境中多人生理状态的完整解决方案。*我们建议使用雷达系统对人们进行远程非接触式定位，对他们的活动进行分类，以及估计他们的呼吸频率。雷达系统将增加新开发的无源射频识别(RFID)系统，该系统将有助于定位和区分多人。在跟踪和监视单个人的情况下，仅使用雷达、非接触式解决方案就足够了。*这项研究将把重点放在开发新的算法，基于雷达和RFID系统的信号来定位、估计呼吸频率、分类和学习人的活动。将开发用于确定到达方向的RFID读取器的新硬件和算法。这些算法将是健壮的，并被设计成在现实环境中运行。*五名研究生和两名博士后研究员将在这一5年期计划中接受培训。该项目对加拿大很重要，因为它将改善加拿大和国外人民的生活质量。此外，它还有可能显著节省医疗成本，因为它将使医务人员能够发现正在进行的事件，如老年人摔倒，并及时做出反应。**

The long-term objective of the proposed research program is to develop a system for comprehensive detection of activities as well as physiological assessment of people who need assistance in their natural environments. For example, this includes tracking the daily activities of elderly people, people with dementia, children, inmates, as well as detecting falls of elderly people and sudden life-threatening events such as heart attack. This project will allow for continuous monitoring of a person who lives alone or with other people, or for simultaneous monitoring of several people. The approach taken in this NSERC Discovery research program differs from research based on smart home and recognition of daily activities. Here, a complete solution for tracking and assessment of physiological states of multiple people in realistic environments based only on radio-frequency backscattering signals is proposed. *** We propose using a radar system for remote contactless localization of people, classification of their activities as well as estimation of their breathing rates. The radar system will be augmented with newly developed passive radio frequency identification (RFID) system that will assist in localization and distinguishing among multiple people. In cases when a single person is tracked and monitored, radar-only non-contact-based solution is sufficient. *** This research will place a strong emphasis on the development of new algorithms for localization, breathing rate estimation, classification and learning about the activities of people based on signals from radar and RFID systems. New hardware and algorithms for the RFID reader for determining direction of arrivals will be developed. The algorithms will be robust and designed to operate in realistic environments. *** Five graduate students and two post-doctoral fellows will be trained in this 5-year program. The project is important for Canada because it will result in improving quality of life of the people of Canada and abroad. In addition, it has a potential to result in significant savings in health-care costs because it will allow medical workers to detect events in progress, such as falls of elderly people, and to react on time. **