Datasets:

Project-AgML
/

banana_guava_quality_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': Class_A
            '1': Class_B
            '2': Defect
    - name: crop_type
      dtype: string
  splits:
    - name: train
      num_bytes: 467881435
      num_examples: 1748
  download_size: 477334142
  dataset_size: 467881435

Banana Guava Quality Classification

A dataset for quality classification of bananas and guavas. The dataset contains 1,748 images across 3 classes: Class_A, Class_B, Defect.
Images per class:

Class_A: 671
Class_B: 469
Defect: 608

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

Citation

@article{kumari2024banana,
  title={Banana and Guava dataset for machine learning and deep learning-based quality classification},
  author={Kumari, Abiban and Singh, Jaswinder},
  journal={Data in Brief},
  volume={57},
  pages={111025},
  year={2024},
  publisher={Elsevier}
}

KUMARI, ABIBAN; Singh, Jaswinder (2024), “Fruits (Banana and Guava) datasets for non-destructive quality classifications”, Mendeley Data, V2, doi: 10.17632/56td5w4wz2.2