Massive Trajectory Matching and Construction from Aerial Videos based on Frame-by-Frame Vehicle Detections

TL;DR

This paper introduces a novel framework for constructing massive vehicle trajectories from UAV aerial videos, utilizing CNN-based detection, dynamic feature matching, and denoising techniques, achieving high accuracy in congested and free-flow traffic conditions.

Contribution

It presents a new framework combining vehicle detection, dynamic feature-based matching, and denoising for large-scale trajectory construction from aerial videos.

Findings

Achieved over 93% recall and 98% precision in trajectory extraction.

Effective in both congested and free-flow traffic conditions.

Processing speed of approximately 30 seconds per trajectory.

Abstract

Vehicle trajectory data provides critical information for traffic flow modeling and analysis. Unmanned aerial vehicles (UAV) is an emerging technology for traffic data collection because of its flexibility and diversity on spatial and temporal coverage. Vehicle trajectories are constructed from frame-by-frame detections. The increase of vehicle counts makes multiple-target matching more challenging. Errors are caused by pixel jitter, vehicle shadows, road marks as well as some missing detections. This research proposes a novel framework for construction of massive vehicle trajectories from aerial videos by matching vehicle detections based on traffic flow dynamic features. The You Look Only Once (YOLO) v4 is used for vehicle detection in UAV videos based on Convolution Neural Network (CNN). Trajectory construction is proposed in detected bounding boxes with trajectory identification, integrity enhancement, and coordinate transformation from image coordinates to the Frenet coordinates. The raw trajectory obtained is then denoised by the ensemble empirical mode decomposition (EEMD). Our framework is tested on two aerial videos taken by a UAV on city expressway covering congested and free-flow traffic conditions. The results show that the proposed framework achieves a Recall of 93.00% and 86.69%, and a Precision of 98.86% and 98.83% for vehicle trajectories in the free-flow and congested traffic conditions.The trajectory processing speed is about 30s per track.