Evaluating Fastest Path Procedures on Roundabouts by Extracting Vehicle Trajectories from Unmanned Aerial Vehicles

Abstract

Unmanned aerial vehicles are one of the most emerging technologies being used recently in transportation field. The collection and analysis of high-resolution data such as vehicles’ trajectories are challenging concerning the analysis of driving behavior for use on optimal road design. At present, the expected speeds of vehicles on roundabouts are calculated based on fastest path analysis techniques that several guidelines provide. This paper presents the results of experimental research be conducted as a first step in developing a model for the prediction of vehicle paths on roundabouts. The research aims to define the basic path elements of vehicles maneuvers at multilane roundabouts on free-flow conditions. An analytical guiding framework regarding the use of UAVs for data acquisition in traffic analysis is presented. The proposed methodology follows a simple and low-cost way of extracting detailed naturalistic vehicle trajectories data through UAVs. Video image processing techniques and geographic information system methods are used and provide a useful tool for obtaining spatiotemporal phenomena at two multi-lane roundabouts. The level of accuracy of the method is evaluated by minimizing the errors due to camera instability and distortion effects. The results of the experiment highlight the usefulness of the proposed UAV and video processing technique for further driving behavior analysis. The analysis that was conducted indicates differences between observed vehicles trajectories and existing fastest paths procedures.