Project-AgML/wheat_growth_stage_classification

Wheat Growth Stage Classification

This dataset provides real RGB images of winter wheat plants at various growth stages, captured in a controlled laboratory environment. The images were collected over a 15-month period from April 2021 to June 2022 using a fixed imaging platform, ensuring consistent visual representation of wheat development under stable conditions. The dataset contains 450 images across 15 classes: FM-ED, FM-LD, FM-MD, FM-OM, FM-SD, HF-ED, HF-LD, HF-MD, HF-OM, HF-SD, RJ-ED, RJ-LD, RJ-MD, RJ-OM, RJ-SD.
Images per class:

• FM-ED: 30
• FM-LD: 30
• FM-MD: 30
• FM-OM: 30
• FM-SD: 30
• HF-ED: 30
• HF-LD: 30
• HF-MD: 30
• HF-OM: 30
• HF-SD: 30
• RJ-ED: 30
• RJ-LD: 30
• RJ-MD: 30
• RJ-OM: 30
• RJ-SD: 30

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{obsie2025enhancing,
  title={Enhancing drought prediction through machine learning: advanced techniques combining phenotypic and agrometeorological data},
  author={Obsie, Efrem Yohannes and Liu, Yongguo},
  journal={Smart Agricultural Technology},
  volume={12},
  pages={101227},
  year={2025},
  publisher={Elsevier}
}

https://www.kaggle.com/datasets/jianbinyao/minimum-dataset/data

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