Project-AgML/ws_dino_weed_segmentation

Ws Dino Weed Segmentation

This dataset provides real RGB aerial imagery captured from UAVs over agricultural fields in southern Germany, specifically focusing on sugar beet and sorghum crops. The images were collected using DJI UAV platforms equipped with high-resolution RGB cameras, offering detailed field-level views for weed segmentation tasks. It serves as a practical resource for developing computer vision models in precision agriculture applications. The dataset contains 19 images with pixel-level mask annotations.

This dataset is indexed on https://project-agml.github.io/ as part of the AgML python library.

The original train/test/val split has been preserved in the split column.

Citation

@article{zhou2026ws,
  title={WS-DINO: A DINOv2-Based Weed Segmentation Method with Feature Priors and Spatial Fusion},
  author={Zhou, Hongsheng and Liu, Jiangping and Wu, Rigeng and Zhao, Baoping},
  journal={Agriculture},
  volume={16},
  pages={1105},
  year={2026},
  publisher={MDPI}
}

This dataset was reformatted from its original format to match HuggingFace standards.