Datasets:

Project-AgML
/

tropical_flower_variety_classification

configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-4.0
task_categories:
  - image-classification
size_categories:
  - 1K<n<10K
dataset_info:
  features:
    - name: image
      dtype: image
    - name: label
      dtype:
        class_label:
          names:
            '0': Bougainvillea
            '1': Crown of thorns
            '2': Hibiscus
            '3': Jungle geranium
            '4': Madagascar periwinkle
            '5': Marigold
            '6': Rose
  splits:
    - name: train
      num_bytes: 1712020186
      num_examples: 4319
  download_size: 1666159806
  dataset_size: 1712020186

Tropical Flower Variety Classification

A dataset for variety classification of tropical flowers. The dataset contains 4,319 images across 7 classes: Bougainvillea, Crown of thorns, Hibiscus, Jungle geranium, Madagascar periwinkle, Marigold, Rose.
Images per class:

Bougainvillea: 580
Crown of thorns: 583
Hibiscus: 548
Jungle geranium: 698
Madagascar periwinkle: 366
Marigold: 717
Rose: 827

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

Citation

@article{rahat2025dataset,
  title={Dataset of seven tropical flower species from Bangladesh: A resource for classification and ecological studies},
  author={Rahat, Riazul Islam and Hossain, Md Sohag and Mojumdar, Mayen Uddin and Chakraborty, Narayan Ranjan},
  journal={Data in Brief},
  volume={59},
  pages={111374},
  year={2025},
  publisher={Elsevier}
}

Rahat, Riazul Islam; hossain, Md.Sohag; Mojumdar, Mayen Uddin ; Chakraborty, Narayan Ranjan; Noori, Sheak Rashed Haider; Siddiquee, Shah Md Tanvir (2024), “Tropical Flower Dataset: Seven Species from Bangladesh for Classification and Ecological Research.”, Mendeley Data, V1, doi: 10.17632/njfg9nh92t.1