Assessment of modern smartphone sensors performance on vehicle localization in urban environments

Theodoros Lazarou; Chris Danezis

doi:10.1117/12.2292354

Translator Disclaimer

6 September 2017 Assessment of modern smartphone sensors performance on vehicle localization in urban environments

Theodoros Lazarou, Chris Danezis

Author Affiliations +

Proceedings Volume 10444, Fifth International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2017); 104441S (2017) https://doi.org/10.1117/12.2292354
Event: Fifth International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2017), 2017, Paphos, Cyprus

Abstract

The advent of Global Navigation Satellite Systems (GNSS) initiated a revolution in Positioning, Navigation and Timing (PNT) applications. Besides the enormous impact on geospatial data acquisition and reality capture, satellite navigation has penetrated everyday life, a fact which is proved by the increasing degree of human reliance on GNSS-enabled smart devices to perform casual activities. Nevertheless, GNSS does not perform well in all cases. Specifically, in GNSS-challenging environments, such as urban canyons or forested areas, navigation performance may be significantly degraded or even nullified. Consequently, positioning is achieved by combining GNSS with additional heterogeneous information or sensors, such as inertial sensors. To date, most smartphones are equipped with at least accelerometers and gyroscopes, besides GNSS chipsets. In the frame of this research, difficult localization scenarios were investigated to assess the performance of these low-cost inertial sensors with respect to higher grade GNSS and IMU systems. Four state-of-the-art smartphones were mounted on a specifically designed on-purpose build platform along with reference equipment. The platform was installed on top of a vehicle, which was driven by a predefined trajectory that included several GNSS-challenging parts. Consequently, positioning and inertial readings were acquired by smartphones and compared to the information collected by the reference equipment. The results indicated that although the smartphone GNSS receivers have increased sensitivity, they were unable to produce an acceptable solution for more than 30% of the driven course. However, all smartphones managed to identify, up to a satisfactory degree, distinct driving features, such as curves or bumps.

Citation Download Citation

Theodoros Lazarou and Chris Danezis "Assessment of modern smartphone sensors performance on vehicle localization in urban environments", Proc. SPIE 10444, Fifth International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2017), 104441S (6 September 2017); https://doi.org/10.1117/12.2292354

ACCESS THE FULL ARTICLE