edouard-rolland/zebra-herds-aerial

Dataset Description

This dataset presents labelled top-view (nadir) aerial images of plains zebra (Equus quagga) herds. The footage was captured with a DJI Mini-series drone flying at an altitude of 60 metres, in January 2025 during a field campaign at the Ol Pejeta Conservancy, Laikipia County, Kenya. Individual video frames were extracted and every visible zebra was annotated with a bounding box, making the dataset suitable for training and evaluating object-detection models for the automated detection and counting of zebras in drone imagery.

The data was collected as part of the WildDrone project, which investigates the use of autonomous drones and drone swarms for non-intrusive monitoring of large mammals in their natural habitats.

Annotations are provided as COCO-style bounding boxes [x, y, width, height] in absolute pixel coordinates, stored under the objects field of each example.

Citation

@InProceedings{10.1007/978-3-032-07638-0_22,
  author    = {Rolland, Edouard G. A. and Meier, Kilian and Gr{\o}ntved, Kasper A. R. and Laporte-Devylder, Lucie and Maalouf, Guy and Lundquist, Ulrik P. S. and Christensen, Anders L.},
  editor    = {Mathieu, Philippe and De la Prieta, Fernando},
  title     = {Drone Swarms for Multi-perspective Monitoring of Large Mammals in their Natural Habitats: Deployment and Field Trials},
  booktitle = {Advances in Practical Applications of Agents, Multi-Agent Systems, and Computational Social Science: The PAAMS Collection},
  year      = {2025},
  publisher = {Springer Nature Switzerland},
  address   = {Cham},
  pages     = {266--277},
  isbn      = {978-3-032-07638-0}
}

Acknowledgement

This work is supported by the WildDrone MSCA Doctoral Network funded by EU Horizon Europe under grant agreement no. 101071224, and by the Innovation Fund.

Dataset Labels

['zebra']

Example of Labelled Images

Annotated Zebras

The image below is a patchwork of every annotated zebra in the dataset — each tile is a single bounding-box crop. It illustrates the scale of the 6,159 annotations and the variety of poses, herd densities, and backgrounds captured from a top-view drone perspective.

Number of Images

{'train': 424, 'validation': 107}

Intended Uses

The dataset is intended for training and evaluating object-detection models (for example the YOLO family) for the automated detection and counting of zebras in top-view drone imagery. Such models support non-intrusive wildlife monitoring and conservation efforts. More context on the drone-swarm monitoring system behind this data collection can be found in the paper listed in the Citation section.

How to Use

• Install datasets:

pip install datasets

• Load the dataset:

from datasets import load_dataset

ds = load_dataset("edouard-rolland/zebra-herds-aerial")
example = ds['train'][0]

License

MIT