Life-logging of wheelchair driving on web maps for visualizing potential accidents and incidents

  • Authors:
  • Yusuke Iwasawa;Ikuko Eguchi Yairi

  • Affiliations:
  • Graduate School of Science and Technology, Sophia University, Chiyodaku, Tokyo, Japan;Graduate School of Science and Technology, Sophia University, Chiyodaku, Tokyo, Japan

  • Venue:
  • PRICAI'12 Proceedings of the 12th Pacific Rim international conference on Trends in Artificial Intelligence
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Life-logging has attracted rising attention as the most fundamental elements for developing every rich software today. This paper presents computational estimation and mapping of potential accidents and incidents of wheelchairs from life-logs with a single cheap and mini-sized three-axis accelerometer mounted on a wheelchair. Wheelchair driving data was obtained by real wheelchair users driving with their wheelchair on real roads, but has the sampling time delay and noises. As a first step of computational estimation, wheelchair driving behavior was classified into moving and static action, and the moving action was divided into tough and smooth status of the ground surface. We employed Support Vector Machine for classification, and made the precise supervised data from the video of wheelchair driving. As the result of classification, estimation of moving/static was achieved 98.2% accuracy rate and estimation of tough/smooth surface was achieved 82.6% accuracy rate. From the surface estimation result, wheelchair-driving difficulty was mapped and evaluated.