A Self-dependent Camera/Map-Aided Smartphone-Based PNS

To build a portable and reliable indoor pedestrian navigation system (PNS), signal availability, cost, power consumption, and accuracy are the key issues need to be considered and evaluated. Taken all the above issues into consideration, a totally self-dependent and low-cost camera/map-aided smartphone-based indoor pedestrian navigation system is proposed in this research. Different from the traditional indoor navigation system, in this research, only smartphone-embedded sensors, i.e. Micro-electromechanical System (MEMS) gyroscope, MEMS accelerometers, magnetometer and camera, are needed to establish the positioning system. To effectively use the indoor geospatial information, a wall-cross method is proposed to add indoor map information to the positioning system through a novel cascade structure algorithm. Moreover, a low-cost line segment detection (LSD) - based map acquisition method is utilized to obtain the indoor map information. Nonlinear particle filter is applied to implement the map aiding process to the navigation solutions. The proposed method has been tested and verified through real experiments in different scenarios. This research has important theoretical and practical value for both industry and academic areas.