Title: Motion mode recognition in multi-storey buildings based on the naive Bayes method

Authors: Litao Han; Lei Gu; Cheng Gong; Tianfa Wang; Aiguo Zhang

Addresses: College of Geodesy and Geomatics, Shandong University of Science and Technology, Qingdao, 266590, China ' College of Geodesy and Geomatics, Shandong University of Science and Technology, Qingdao, 266590, China ' College of Geodesy and Geomatics, Shandong University of Science and Technology, Qingdao, 266590, China ' College of Geodesy and Geomatics, Shandong University of Science and Technology, Qingdao, 266590, China ' College of Computer and Information Engineering, Xiamen Institute of Technology, Xiamen, 361024, China

Abstract: When walking within a multi-storey building, pedestrians use a variety of modes of motion, such as going upstairs or downstairs, or walking on a plane. In each of these modes, the step size will be different, and this has a strong impact on the accuracy of pedestrian dead reckoning. In order to identify the patterns of movement in multi-storey buildings, we propose a new method of movement recognition based on smart phones. Firstly, the relationships between the patterns of indoor movement and the changes in the air pressure, acceleration and angular velocity data obtained from the built-in sensors of mobile phones are analysed. The naive Bayes method is then used to identify four different modes of motion: walking up and down the stairs, moving on a landing and moving at a constant speed along a corridor. Our experimental results show that the recognition accuracy of our scheme reaches 95.82%.

Keywords: accelerometer; barometer; gyroscope; motion mode recognition; naive Bayes; smart phone.

DOI: 10.1504/IJSNET.2022.10045945

International Journal of Sensor Networks, 2022 Vol.38 No.3, pp.166 - 176

Received: 25 May 2021
Accepted: 01 Jun 2021

Published online: 24 Mar 2022 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article