Detection of multiple human location and direction by integrating different kinds of sensors

  • Authors:
  • Tsuyoshi Shinno;Kazunori Hashizume;Junichi Tajima;Shigeo Kaneda;Hirohide Haga

  • Affiliations:
  • Doshisha University, Miyakotani Kyotanabe, Japan;Doshisha University, Miyakotani Kyotanabe, Japan;Doshisha University, Miyakotani Kyotanabe, Japan;Doshisha University, Miyakotani Kyotanabe, Japan;Doshisha University, Miyakotani Kyotanabe, Japan

  • Venue:
  • Proceedings of the 1st international conference on PErvasive Technologies Related to Assistive Environments
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

As ubiquitous services become available, information on human location and direction have so important that many detection methods have been proposed. However, a lot of them generate electromagnetic waves or light rays. They are unsuited for use in detection of children's location because parents are wary of health damage caused by radiated electromagnetic waves. The accuracy of a lot of methods which doesn't generate electromagnetic waves is not always high indoors. In this paper, we propose a detection method of indoor multiple human location and direction by using a stereovision camera and direction detection sensors. With a stereovision camera, it is possible to detect human location with high accuracy, but individuals cannot be identified. Direction detection sensors make it possible to identify an individual wearing them. By integrating these sensing devices based on direction information, detection of location and direction become possible with high accuracy. Moreover, with this method it is possible to identify an individual. We built a prototype and conducted an evaluation experiment. We conducted the experiment in a room divided by a ferroconcrete wall. We conducted a total of 10 experiments, changing the number of subjects. The experimental evaluation result shows the average error of location was 0.11[m] and the average error of direction was 0.39[rad]. The success rate of identifying an individual was 90[%].