Human Localization by acceleration sensors located on the floor in a home environment

Abstract

Recent advances in information communication technologies have yielded a variety of new services. Location based services are attractive as they offer so many new functionalities. However, if they are used indoors, location recognition is essential because the global positioning system, the preferred choice outdoors, cannot be used indoors. Therefore, this paper proposes an indoor human location recognition method that uses the vibration signals captured by acceleration sensors placed on the floor. The user does not wear any device. Preliminary experiments show that signal magnitude corresponds to the distance between the sensor and the point of pressure which indicates the feasibility of the proposal. Another experiment with walking subjects showed that for subject-sensor distances of 1 to 5 m, the proposed method could recognize subject location with 68 % accuracy at distances of 1 m or better tolerance and 87 % at distances of 2 m or better tolerance. These results verify the feasibility of the proposed system.