Project-AgML/MH_Weed16_weed_variety_classification

MH Weed16 Weed Variety Classification

A dataset for variety classification of weeds. The dataset contains 19,141 images across 16 classes:
Images per class:

• Asian_Pigeonwings_(Clitoria Ternatea): 678
• Bilayat_(Mexicana_Argemone): 859
• Choti_dudhi_(Euphorbia_hirta): 580
• Digitaria_SP_(Digitaria Sanguinalis ): 1,561
• Dwarf_cassia_(Chamaecrista pumila): 1,306
• Gajar_gavat_(Parthenium hysterophorus): 1,543
• Graceful_Sandmart_(Euphorbia hypericifolia): 262
• Harali_(Cynodon_dactylon): 709
• Kena_(Commplina_benghalensio): 1,704
• Lamber_Quarter_plant(Chenopodium ): 836
• Lavhala_(Cyperus_Rotundus): 1,177
• Little_Mallow(Malva parviflora): 2,100
• Moti_dudhi(Euphorbia_geneculata_L): 3,002
• Obscure_morning _glory(Ipomoea obscura): 1,637
• Punarnava _(Boerhaavia diffusa): 544
• Sicklepod_(Senna obtusifolia): 643

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

Citation

@article{shinde2025indian,
  title={An Indian annotated weed dataset for computer vision tasks in precision farming},
  author={Shinde, Sayali and Attar, Vahida},
  journal={Data in Brief},
  volume={61},
  pages={111691},
  year={2025},
  publisher={Elsevier}
}

Shinde, Sayali; Attar, Dr. Vahida; Technological University Pune, COEP; Technology Innovation Hub, Indian Statistical Institute Kolkata, IDEAS (2025), “MH-Weed16:An Indian Multiclass Annotated Weed Dataset for Computer Vision Tasks ”, Mendeley Data, V2, doi: 10.17632/d3n3mgjjbv.2