Skip to content

Latest commit

 

History

History
42 lines (34 loc) · 1.95 KB

File metadata and controls

42 lines (34 loc) · 1.95 KB

DroneDetectionDataset (GitHub)

Overview

The Drone Detection Dataset is a real-world object detection dataset specifically created for Quadcopter Unmanned Aerial Vehicle (UAV) detection. It was developed as part of a PhD research on Engineering Applications of Artificial Intelligence at Warsaw University of Technology.

Dataset Statistics

  • 51,446 training images
  • 5,375 test images
  • All images are 640x480 RGB format
  • Shows drones in different types, sizes, scales, positions, environments, and times-of-day
  • Includes XML labels prepared for Haar Cascade training (easily modified for ANN training)
  • Total dataset size is approximately 5GB

Dataset Creation

  • Frames were automatically extracted from publicly available YouTube videos
  • Manually annotated by the author
  • Initial labeling was done manually, with subsequent images initially labeled by trained models to reduce annotation time
  • Frames were extracted at approximately 50-140 frame intervals

Technical Details

  • Includes trained Haar Cascades and ANN models
  • 10 best models (sorted by accuracy) referenced in the original paper are available
  • Dataset focuses on drone detection rather than tracking
  • Designed to avoid model overfitting by including diverse environments

Characteristics for Drone Deployment

  • Contains diverse drone types and environments
  • Includes different lighting conditions and times of day
  • Specifically designed for security applications like protecting critical infrastructure
  • Suitable for both traditional computer vision (Haar Cascades) and deep learning approaches

License

  • Released under MIT License for image tags
  • Video sources are from publicly available YouTube videos

Authors

  • Maciej Pawełczyk
  • Marek Wojtyra

Source

GitHub: https://github.com/Maciullo/DroneDetectionDataset Paper: M. Ł. Pawełczyk and M. Wojtyra, "Real World Object Detection Dataset for Quadcopter Unmanned Aerial Vehicle Detection," in IEEE Access, vol. 8, pp. 174394-174409, 2020