Vehicle Trajectory Dataset from Drone Videos Including Off-Ramp and Congested Traffic
Vehicle trajectory data have become essential for many research fields, such as traffic flow, traffic safety and automated driving. In order to make trajectory data usable for researchers, an overview of the included road section and traffic situation as well as a description of the data processing methodology is necessary. In this paper, we present a trajectory dataset from a German highway with two lanes per direction, an off-ramp and congested traffic in one direction, and an on-ramp in the other direction. The dataset contains 8,648 trajectories and covers 87 minutes and a ~1,200 m long section of the road. The trajectories were extracted from drone videos using a post-trained yolov5 object detection model and projected onto the road surface using a 3D camera calibration. The post-processing methodology can compensate for most false detections and yield accurate speeds and accelerations. We present some applications of the data including a traffic flow analysis and accident risk analysis. The trajectory data are also compared with induction loop data and vehicle-based smartphone sensor data in order to evaluate the plausibility and quality of the trajectory data. The deviations of the speeds and accelerations are estimated at 0.45 m/s and 0.3 m/s2 respectively.
| Location | Highway A43 near Münster, Germany 2 lanes per direction Two-lane off-ramp (West to East), one-lane on-ramp (East to West) ~1200 m |
| Time | 06th September 2021 07:11 a.m. to 08:38 a.m. (87 min) |
| Number of trajectories | 5016 (West to East) 3632 (East to West) |
When using data from this dataset, please cite the dataset as follows:
Berghaus, M., Lamberty, S., Ehlers, J., Kallo, E., Oeser, M. (2024). Vehicle Trajectory Dataset from Drone Videos Including Off-Ramp and Congested Traffic. Communications in Transportation Research, 4.
