Indoor Pedestrian Dead Reckoning Calibration by Visual Tracking and Map Information

Abstract

Currently, Pedestrian Dead Reckoning (PDR) systems are becoming more attractive in market of indoor positioning. This is mainly due to the development of cheap and light Micro Electro-Mechanical Systems (MEMS) on smartphones and less requirement of additional infrastructures in indoor areas. However, it still faces the problem of drift accumulation and needs the support from external positioning systems. Vision-aided inertial navigation, as one possible solution to that problem, has become very popular in indoor localization with satisfied performance than individual PDR system. In the literature however, previous studies use fixed platform and the visual tracking uses feature-extraction-based methods. This paper instead contributes a distributed implementation of positioning system and uses deep learning for visual tracking. Meanwhile, as both inertial navigation and optical system can only provide relative positioning information, this paper contributes a method to integrate digital map with real geographical coordinates to supply absolute location. This hybrid system has been tested on two common operation systems of smartphones as iOS and Android, based on corresponded data collection apps respectively, in order to test the robustness of method. It also uses two different ways for calibration, by time synchronization of positions and heading calibration based on time steps. According to the results, localization information collected from both operation systems has been significantly improved after integrating with visual tracking data.

3-D Passive-Vision-Aided Pedestrian Dead Reckoning for Indoor Positioning

“…With these improvements, the 2D accuracy (0.16m) of this system is significantly higher than the best performances provided by the Commercial Mobile Radio Service (CRMS) reported in FCC (5~10 m) [54]. Moreover, the previous studies [11,[22][23][24][51][52][53] only provide 2D user locations, while to enable a continuous positioning service, particularly for the time the user is walking up or downstairs, a 3D (or the recognition of the floors) are required [55][56][57].…”

“…During the operation, the smartphone-based PDR keeps actively tracking the user movement, while the OPS only functions in the LoS areas, shifting from one scene to another. This is an update to previous work [51][52][53], as it can handle multi-scene instead of a fixed scene. During the movement, the smartphones are held horizontally and pointing forward.…”

“…However, this approach is not fully practical, particularly for commercial applications, as video recording by the embedded camera is energy consuming and cannot support long durations for indoor localization. Therefore, the authors previously suggested deploying surveillance cameras for pedestrian detection while using inertial sensors in smartphones [51,52].…”

“…This study first improves the design of the previous system in the aspect of 2D PVINS. The video data are used to be derived from a single camera within a fixed scene [51][52][53]. However, in this study, they are acquired from multiple cameras with scene shifting to enlarge the visible areas for continuous tracking of user movement.…”

“…Moreover, the sensor fusion method previously used is based on position replacement by time synchronization, which is not that close to reality. This study employs an alternative approach called heading calibration based on the comparison of the accuracy of these two methods conducted in [52]. It also adds the step length calibration to improve the performance of the PDR system.…”