GPS Trajectory Cleaning For Driving Behaviour Detection System
Abstract— 2017 records of WHO show that road traffic accidents deaths in Myanmar reached 10,527 or 2.67% of total deaths. It also shows that most of the road traffic accidents are due to drivers. Therefore, it is essential to develop a system which is able to detect the drivers’ driving behavior to reduce the traffic accidents. Nevertheless, in order to detect the drivers’ driving behavior, their accurate GPS trajectory is needed and there are still lots of challenges to get accurate GPS trajectory. GPS trajectory cleaning is one of the most important steps for preprocessing GPS data. GPS data might be wrong because sometimes signals transmitted by satellites cannot be recorded accurately by GPS receivers because of the interference, weak signal, and malfunction of sensor. Therefore, the recorded GPS point can be far from the actual location of the receiver. This inaccurate location can make strong affect to some decision making processes based on the GPS data. In order to eliminate wrong GPS data points from the trajectory, some trajectory cleaning processes: detecting stop points, interpolating missing segments and removing inaccurate points are proposed in this paper. Moreover, the results show that preprocessing trajectory cleaning approach helps to improve the quality of trajectory clustering.
Index Terms— Trajectory cleaning, Detecting stop points, Interpolating missing segments, Trajectory clustering
Tin Lai Lai Mon
University of Computer Studies, MYANMAR
Cite: Tin Lai Lai Mon, "GPS Trajectory Cleaning For Driving Behaviour Detection System, " Proceedings of 2020 the 10th International Workshop on Computer Science and Engineering (WCSE 2020), pp. 151-155, Yangon (Rangoon), Myanmar (Burma), February 26-28, 2020.