Datasets:

Project-AgML
/

cottonsim_detection

dataset_info:
  features:
    - name: image
      dtype: image
    - name: objects
      struct:
        - name: bbox
          list:
            list: float64
        - name: categories
          list:
            class_label:
              names:
                '0': class0
    - name: split
      dtype: string
  splits:
    - name: train
      num_bytes: 1224391
      num_examples: 40
  download_size: 1225894
  dataset_size: 1224391
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-4.0
task_categories:
  - object-detection
size_categories:
  - n<1K

Cottonsim Detection

This dataset provides synthetic images of cotton plants in agricultural settings for object detection in crop monitoring. Captured using RealSense RGB-D cameras on a ground-based platform, the imagery simulates field conditions to support vision-guided agricultural robotics development. The dataset contains 40 images with 1,093 bounding box annotations across 1 category.

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{thayananthan2025cottonsim,
  title={CottonSim: A vision-guided autonomous robotic system for cotton harvesting in Gazebo simulation},
  author={Thayananthan, Thevathayarajh and Zhang, Xin and Huang, Yanbo and Chen, Jingdao and Wijewardane, Nuwan K. and Martins, Vitor S. and Chesser, Gary D. and Goodin, Christopher T.},
  journal={Computers and Electronics in Agriculture},
  volume={239},
  pages={110963},
  year={2025},
  publisher={Elsevier}
}

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