Vehicle Trajectory Deviation Data Collection Method Based on Unmanned Aerial Vehicle Aerial Imagery

Traffic risk is higher on curved road sections compared to regular sections. Understanding vehicle trajectory deviation is crucial for analyzing road safety. The conventional method for collecting vehicle trajectory is limited to measuring the deviation of a single vehicle or vehicles on a single lane. To overcome this limitation, this study proposes a novel method that utilizes drone aerial imagery to gather data on vehicle trajectory deviation. By leveraging the bird’s-eye view and high-resolution images provided by drones, we can obtain detailed trajectory information and deviation analysis of multiple vehicles in multiple lanes. The effectiveness and accuracy of this method were confirmed through testing on real road scenarios. The trajectory analysis results reveal significant deviations from the lane lines in tight radius curves for both large and small vehicles. These analysis results provide valuable data for road design, traffic safety assessment, and the development of driving assistance systems.