Project-AgML/seed_germination_detection

Seed Germination Detection

Time-series images of germination experiments for maize (Zea mays), rye (Secale cereale) and pearl millet (Pennisetum glaucum): 2,449 seeds on black cloth in petri dishes, photographed repeatedly, giving 23,797 images (624x624, one dish each). Every seed is boxed and labelled germinated (radicle emerged from the seed coat) or not germinated. The crop, dish series, time step and the authors' train/validation/test split (by dish series) are kept as columns. The dataset contains 23,797 images with 235,933 bounding box annotations across 2 categories.

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{genze2020accurate,
  title={Accurate machine learning-based germination detection, prediction and quality assessment of three grain crops},
  author={Genze, Nikita and Bharti, Richa and Grieb, Michael and Schultheiss, Sebastian J. and Grimm, Dominik G.},
  journal={Plant Methods},
  volume={16},
  pages={157},
  year={2020},
  publisher={BioMed Central}
}

The dataset itself can be cited as:

Nikita Genze. (2023). Data for: Accurate Machine Learning-Based Germination Detection, Prediction and Quality Assessment of Various Seed Cultivars [Dataset]. Mendeley. https://doi.org/10.17632/4WKT6THGP6.3

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