Pedestrian positioning in urban city with the aid of Google maps street view

Abstract

Pedestrian navigation has become one of the most used services in people's city lives. Not only smartphone based navigation, but also the application in the next generation of intelligent wearable devices, such as smart glasses, attract attentions from both scientists and engineers. The satisfied navigation service requires an accurate positioning technology. Even though the current smartphones have integrated various sensors, such as Global Navigation Satellite System receiver, gyroscope, accelerometer and magnetometer sensors, the performance of positioning in city urban is still not satisfied. The reasons of the errors include GNSS signals reflections, high dynamic of pedestrian activities and disturbance of the magnetic field in city environments. This paper proposes to utilize the camera sensor for improving the accuracy of the positioning. The camera sensor provides the visual observation for surround environment. This observation is compared with the available Google Maps Street View in order to correct positioning errors. With the visual matching between the geo-tagged pedestrian's photo and the reference images from Google Maps Street View, we expect to reduce the positioning error into 4 meters, and further recognize which side of the road or which corner of the crossroads the pedestrian is in.