The Swarm Intelligence Freeway-Urban Trajectories (SWIFTraj) Dataset - Part II: A Graph-Based Approach for Trajectory Connection

TL;DR

Results from real-world experiments show the effectiveness of the proposed method in addressing key challenges in UAV-based trajectory connection and highlight its potential for large-scale vehicle trajectory collection.

Abstract

In Part I of this companion paper series, we introduced SWIFTraj, a new open-source vehicle trajectory dataset collected using a unmanned aerial vehicle (UAV) swarm. The dataset has two distinctive features. First, by connecting trajectories across consecutive UAV videos, it provides long-distance continuous trajectories, with the longest exceeding 4.5 km. Second, it covers an integrated traffic network consisting of both freeways and their connected urban roads. Obtaining such long-distance continuous trajectories from a UAV swarm is challenging, due to the need for accurate time alignment across multiple videos and the irregular spatial distribution of UAVs. To address these challenges, this paper proposes a novel graph-based approach for connecting vehicle trajectories captured by a UAV swarm. An undirected graph is constructed to represent flexible UAV layouts, and an automatic time alignment method based on trajectory matching cost minimization is developed to estimate optimal time offsets across videos. To associate trajectories of the same vehicle observed in different videos, a vehicle matching table is established using the Hungarian algorithm. The proposed approach is evaluated using both simulated and real-world data. Results from real-world experiments show that the time alignment error is within three video frames, corresponding to approximately 0.1 s, and that the vehicle matching achieves an F1-score of about 0.99. These results demonstrate the effectiveness of the proposed method in addressing key challenges in UAV-based trajectory connection and highlight its potential for large-scale vehicle trajectory collection.

Figures (11)

• Figure 1: Example of video graph. (a) (Left) The flight plan of pNEUMA dataset (the image is from barmpounakis2020new). (Right) The video graph of pNEUMA dataset. (b) (Left) The flight plan of MAGIC dataset (the image is from ma2022magic). (Right) The video graph of MAGIC dataset.
• Figure 2: The GCVT framework. The subsets of trajectory data from video $v_1$ and $v_2$ are selected to demonstrate the trajectory connection process.
• Figure 3: The spatial transformation of images. The four orange points are the selected feature points.
• Figure 4: Study area and the video graph. (Up) Satellite map of the study area and the recording area of each UAV. (Down) Video graph for the UAV swarm
• Figure 5: Variation of the mean trajectory matching cost defined in Eq. \ref{['eq:trajectory_difference']} for different alignment edges during the search process. The blue curve shows the cost evolution under the bidirectional search, while the orange dashed curve indicates the truncated cost trajectory after the early stopping criterion is triggered.