CAFOSat: CAFO Infrastructure Dataset

CAFOSat is a remote sensing dataset designed for identifying and classifying Concentrated Animal Feeding Operations (CAFOs) across various U.S. states. It includes high-resolution image patches, infrastructure annotations, bounding boxes, and experimental train-test splits for multiple configurations.

🔗 Resources

• GitHub Repository: oishee-hoque/CAFOSat
• Explore the Dataset: CAFOSat Data Loader and Examples

Dataset Structure

The dataset is organized into extracted folders originally packaged as .tar.gz files:

• STATE_filtered/: Original patches per U.S. state
  (e.g., IA_filtered/, AL_filtered/)
• negative_samples/: Verified non-CAFO examples
• barn/, manure_pond/, others/: Augmented synthetic patches by infrastructure type
• CAFOSat.csv: Master metadata file with labels, bounding boxes, and split flags

All image paths referenced in the CSV point to these extracted folders. 📄 Example data loader available here.

---

🔖 Image File Reference (patch_file)

Each row in the metadata includes a patch_file field that provides the relative path to the associated image file.

Example:
-IA_filtered.tar.gz::IA_filtered/crop_4517_patch_10147_Swine_Nursery_IA.tif

• barn.tar.gz::barn/aug_patch_00123.tif`

• negative_sample.tar.gz::barn/neg_patch_00098.tif`

---

Features (CAFOSat.csv)

Column	Description
patch_file	Path to the image file
label	Integer label for class (0–6)
barn, manure_pond, grazing_area, others	Binary flags indicating infrastructure types
geom_bbox	Bounding box coordinates [x1, y1, x2, y2]
geometry	Geospatial polygon outlining the CAFO region
poly_crs	Coordinate reference system used for geometry
patch_crs	Coordinate reference system used for the image patch
category	CAFO class name (e.g., Swine, Dairy)
state	U.S. state where the patch is located
verified_label	Boolean indicating if the label is human-verified
CAFO_UNIQUE_ID	Unique identifier for each CAFO facility
image_type	Image variant type: augmented else real
orig_patch_file	Name/path of the original patch (applicable if the image_type is augmented)
prompt	Text prompt or description for generative/semantic use (if applicable)
weak_x, weak_y	Collected CAFO center coordinates
refined_x, refined_y	Refined CAFO center coordinates
patch_res	Spatial resolution of the patch in meters per pixel
split columns	Flags for different train/test/val splits

---

🧪 Train/Test/Val Split Flags

These are boolean-like flags (True/False) that indicate inclusion in various dataset splits:

• cafosat_verified_training_train, cafosat_verified_training_test, cafosat_verified_training_val
• cafosat_all_training_*
• cafosat_training_set1_*, set2_*
• cafosat_merged_training_*
• cafosat_augmented_training_*

Each flag is a binary indicator (1 = in split, 0 = excluded).

---

Labels

Class ID	Class Name
0	Negative
1	Swine
2	Dairy
3	Beef
4	Poultry
5	Horses
6	Sheep/Goats

---

📌 Intended Use

• CAFO detection and classification – Identify and categorize CAFO facilities from aerial imagery.
• CAFO area detection with bounding box – Localize CAFOs using annotated bounding box coordinates.
• Agricultural infrastructure mapping – Map features like barns, manure ponds, and grazing areas.
• Weak supervision and semi-supervised learning – Leverage partially labeled data for model development.
• Remote sensing benchmark development – Support standardized evaluation for remote sensing tasks.

---

Citation

*TBD

---

License

This dataset is released under the CC BY 4.0 License.

You are free to share, use, and adapt the data with attribution.

Contact

For questions or contributions, contact gza5dr@virginia.edu.