FBIS-22M

Field Boundary Instance Segmentation - 22M dataset (FBIS-22M) - large-scale, multi-resolution dataset comprising 672,909 high-resolution satellite image patches (0.25 m – 10 m) and 22,926,427 instance masks of individual fields.
Note: We have excluded images from the Pleiades satellite mission due to third-party usage policies. This does not impact the dataset’s coverage or quality.

---

Dataset Structure

The dataset is split into multiple archive parts. Additional files:

File	Description
train.txt	List of training image filenames
test.txt	List of testing image filenames

---

How to Use

Download and Recombine

1. Download all .zip.part-* files.
2. Recombine them into a single .zip archive:

Linux / macOS

cat FBIS-22M.zip.part-* > FBIS-22M.zip

Windows

Use 7-Zip → Tools > Combine Files (select all parts).

Extract

unzip FBIS-22M.zip

Citation

If you find our dataset useful in your research, please consider citing it:

@article{lavreniuk2025delineateanything,
      title={Delineate Anything: Resolution-Agnostic Field Boundary Delineation on Satellite Imagery}, 
      author={Mykola Lavreniuk and Nataliia Kussul and Andrii Shelestov and Bohdan Yailymov and Yevhenii Salii and Volodymyr Kuzin and Zoltan Szantoi},
      year={2025},
      journal={arXiv preprint arXiv:2504.02534},
}