Dataset: Traffic Images Captured from UAVs for Use in Training Machine Vision Algorithms for Traffic Management
DOI: 10.3390/data7050053
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper introduces a dataset of Spanish road traffic images captured by unmanned aerial vehicles (UAVs) to address the scarcity of resources for training machine vision algorithms in traffic management. The authors highlight that existing aerial datasets, such as Vis-Drone, are limited by their geographic origin (China), lack of European-specific infrastructure like roundabouts, and language barriers in signage. To support the development of intelligent transport systems, autonomous vehicle cooperation, and traffic violation monitoring, this work provides a specialized dataset focused on complex maneuvers and diverse traffic scenarios. The dataset comprises 15,070 images in PNG format, accompanied by corresponding text files containing annotations in the You Only Look Once (YOLO) format. These annotations identify 155,328 vehicles, categorized into cars (137,602) and motorcycles (17,726). The images were acquired using two UAV models, the DJI Mavic Mini 2 and the Yuneec Typhoon H, equipped with cameras capturing 1920 × 1080 pixel resolution. Data collection adhered to Spanish civilian aviation regulations, with flight heights varying between 35 and 120 meters and camera angles set between 45 and 60 degrees relative to the horizontal axis to capture both the sides and tops of vehicles. The recordings were taken across six distinct locations, including regional roads, urban intersections, rural roads, split roundabouts, and standard roundabouts, ensuring a variety of visibility conditions and vehicle distances to mitigate model overfitting. The authors validated the dataset by training a simple neural network model to ensure data quality and usability. The resulting collection is designed to facilitate the training of convolutional neural networks for object detection in aerial imagery. By providing a robust set of labeled images featuring European traffic infrastructure, particularly roundabouts, the dataset aims to improve the performance of algorithms used for vehicle identification, traffic flow analysis, and emergency response coordination. The authors conclude that this resource fills a critical gap in the field of computer vision for traffic management, offering a practical tool for researchers and developers working on intelligent vehicle systems and UAV-based traffic monitoring.
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-06-20 |
| archive | success | openalex | — | — | 5 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| promote | success | — | — | — | 1 | 2026-06-20 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-20 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Methodological Resource: dataset resource