Datasets:

Project-AgML
/

SDAU_wheat_head_detection

dataset_info:
  features:
    - name: image
      dtype: image
    - name: objects
      struct:
        - name: bbox
          list:
            list: float64
        - name: categories
          list:
            class_label:
              names:
                '0': '0'
    - name: split
      dtype: string
  splits:
    - name: train
      num_bytes: 2244709925
      num_examples: 3030
  download_size: 1784551715
  dataset_size: 2244709925
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-4.0
task_categories:
  - object-detection
size_categories:
  - 1K<n<10K

Sdau Wheat Head Detection

This dataset features synthetic images of wheat heads captured in a laboratory environment. Images were collected using a handheld Canon EOS 70D RGB camera, providing a controlled setting for wheat head detection research. The dataset contains 3,030 images with 96,629 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{sun2025counting,
  title={Counting wheat heads using a simulation model},
  author={Sun, Xiaoyong and Jiang, Tianyou and Hu, Jiming and Song, Zuojie and Ge, Yuheng and Wang, Yongzhen and Liu, Xu and Bing, Jianhao and Li, Jinshan and Zhou, Ziyu and Tang, Zhongzhen and Zhao, Yan and Hao, Jinyu and Zuo, Changzhen and Geng, Xia and Kong, Lingrang},
  journal={Computers and Electronics in Agriculture},
  volume={228},
  pages={109633},
  year={2025},
  publisher={Elsevier}
}

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