An Automated Tool for Detecting and Preventing Unsafe Stair Use by Older Adults
Keywords: Stair monitoring, handrail use, computer vision, motion analysis.
Overview of Research
Accidents on stairs are a leading cause of mortality and morbidity among older adults. The risk of injury posed by stair use is often the deciding factor in allowing older adults to age in their own homes. This project aims to help provide a better understanding of the catalysts of accidents on stairs through the use of state of the art computer vision and machine learning techniques. Currently this work focuses on providing researchers with an automated tool to extract unsafe events on stairs from large databases of video. This is significant as current research is encumbered due to the lack of a comprehensive data set containing unsafe or anomalous events on stairs.
This project will also provide a basis for future systems that will watch people walk down the stairs, analyse their motion in real time, assess if an accident has occurred or will occur, and take appropriate action.
Figure 1: Different methods of image analysis for stair event detection
Jasper Snoek, Alex Mihailidis and Jesse Hoey. (2006). An Automated Tool For Detecting And Preventing Unsafe Stair Use By Older Adults. International Conference on Aging, Disability and Independence. St. Petersburg, FL. Feburary, 2006.
Jasper Snoek, Jesse Hoey, Liam Stewart and Richard Zemel. (2006). Automated Detection and Recognition of Unusual Events on Stairs. In proc. Computer and Robot Vision, Quebec City, QC. June, 2006.
Jasper Snoek, University of Toronto
Alex Mihailidis, University of Toronto
Jesse Hoey, University of Waterloo