Distributed vision-based accident management for assisted living

  • Authors:

  • Hamid Aghajan;Juan Carlos Augusto;Chen Wu;Paul McCullagh;Julie-Ann Walkden

  • Affiliations:

  • Wireless Sensor Networks Lab, Stanford University;School of Computing and Mathematics, University of Ulster, UK;Wireless Sensor Networks Lab, Stanford University;School of Computing and Mathematics, University of Ulster, UK;Ulster Community Hospitals Trust, UK

  • Venue:
  • ICOST'07 Proceedings of the 5th international conference on Smart homes and health telematics
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

We consider the problem of assisting vulnerable people and their carers to reduce the occurrence, and concomitant consequences, of accidents in the home. A wireless sensor network employing multiple sensing and event detection modalities and distributed processing is proposed for smart home monitoring applications. Distributed vision-based analysis is used to detect occupant's posture, and features from multiple cameras are merged through a collaborative reasoning function to determine significant events. The ambient assistance provided will assume minimal expectations on the technology people have to directly interact with. Vision-based technology is coupled with AI-based algorithms in such a way that occupants do not have to wear sensors, other than an unobtrusive identification badge, or learn and remember to use a specific device. In addition the system can assess situations, anticipate problems, produce alerts, advise carers and provide explanations.