Illegal Parking Detection System Based on Scooters' Trajectories

An illegal parking detection system is designed based on scooters' trajectories to identify illegally parked vehicles on the roads. If the trajectories of an area contain more irregular lane switching behavior, the possibility of illegal parking in the area becomes higher. The system has been implemented successfully with the Simulation of Urban MObility (SUMO) simulator. The simulation results show that the system achieves 90.28% of accuracy for detecting illegally parked vehicles.