Project-AgML/CottonWeedID15_classification

Cottonweedid15 Classification

This dataset contains real-world RGB images of 15 common weed species found in cotton agricultural fields, captured under natural field conditions. The images depict weeds growing in typical cotton crop environments, providing a practical resource for developing and evaluating computer vision models for weed identification in precision agriculture applications. The dataset contains 5,187 images across 15 classes: Carpetweeds, Crabgrass, Eclipta, Goosegrass, Morningglory, Nutsedge, PalmerAmaranth, Prickly Sida, Purslane, Ragweed, Sicklepod, SpottedSpurge, SpurredAnoda, Swinecress, Waterhemp.
Images per class:

• Carpetweeds: 763
• Crabgrass: 111
• Eclipta: 254
• Goosegrass: 216
• Morningglory: 1,115
• Nutsedge: 273
• PalmerAmaranth: 689
• Prickly Sida: 129
• Purslane: 450
• Ragweed: 129
• Sicklepod: 240
• SpottedSpurge: 234
• SpurredAnoda: 61
• Swinecress: 72
• Waterhemp: 451

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

Citation

@article{chen2022performance,
  title={Performance evaluation of deep transfer learning on multi-class identification of common weed species in cotton production systems},
  author={Chen, Dong and Lu, Yuzhen and Li, Zhaojian and Young, Sierra},
  journal={Computers and Electronics in Agriculture},
  volume={198},
  pages={107091},
  year={2022},
  publisher={Elsevier}
}

https://www.kaggle.com/datasets/yuzhenlu/cottonweedid15

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