Recursive Bayesian tracking for smart elderly living

摘要

Being able to monitor movements and activities of the elderly in a smart living environment can offer an approach for detecting any on-set of anomalies. These smart living dwellings can be equipped with various networked ambient sensors, wearable sensing technologies and cloud-robotics. Detection of any such on-set of anomalies in elderly movements and activities can further be used to determine the state of mental health of the elderly for example related to dementia or Alzheimer. There are two mains challenges associated with the deployment such sensor network in the dwelling of elderly. The main challenge is in the processing of the sensed information in order to ensure the privacy of individuals. The next big challenge is the robustness of tracking algorithm in the presence of lost or occluded sensed data. It has been shown that the recursive Bayesian approach can offer a suitable framework for tracking targets with the maneuvering trajectories which follows a non-linear behavior and non-Gaussian distribution. This paper presents an overview of recursive Bayesian framework which can be used as a part of tracking environment in the smart living environment of elderly. Through step-by-step development, the paper highlights various features of the method which can be adapted in order to reduce various computational issues associated with the implementation of this framework.