|With the popularity and quick development of smartphones, smartphone based Lifelog that can record users’ actives and locations they stay is in great need. The positioning technique is one of the most important parts in the Lifelog. Different approaches have been proposed to solve the outdoor and indoor positioning issues. Two technologies: Wi-Fi and Pedestrian Dead Reckoning (PDR), could be the compensation or alternative for GPS based outdoor positioning, and also are more promising for indoor positioning. This paper focuses on improving the positioning performance of PDR and Wi-Fi with the aid of mobility context information based map matching. This paper first introduces a method to detect different transportation modes, and then uses transportation mode with other context information like corner and traffic light that have the fixed position to improve the positioning result for the outdoor situation. In addition, this paper extends the idea to the indoor scenario. This research proposes to recognize different vertical activities like taking an elevator/escalator, and use the elevator, escalator or corner as contexts to improve the result of indoor PDR. Moreover, this research uses the improved PDR for automatically creating and updating the Wi-Fi fingerprint radio map in order to improve the performance of the Wi-Fi fingerprint positioning. The experimental results demonstrated that the proposed methods could achieve 3.05-meter positioning accuracy in the outdoor situation and average 2.28-meter positioning accuracy in the indoor situation. Both accurate transportation mode/activity detection results and indoor/outdoor positioning results can improve the quality of a Lifelog application.
Proceedings of the ION 2019 Pacific PNT Meeting
April 8 - 11, 2019
Hilton Waikiki Beach
|540 - 553
|Cite this article:
Gu, Yanlei, Li, Dailin, Kamiya, Yoshihiko, Kamijo, Shunsuke, "Lifelog using Mobility Context Information in Urban City Area," Proceedings of the ION 2019 Pacific PNT Meeting, Honolulu, Hawaii, April 2019, pp. 540-553.
ION Members/Non-Members: 1 Download Credit