---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: objects
    struct:
    - name: bbox
      list:
        list: float64
    - name: categories
      list:
        class_label:
          names:
            '0': grape_bunch
  - name: calibrated
    dtype: bool
  - name: cultivar
    dtype: string
  - name: bbch_stage
    dtype: int32
  - name: capture_date
    dtype: string
  splits:
  - name: train
    num_bytes: 2143586823
    num_examples: 343
  download_size: 2143623277
  dataset_size: 2143586823
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: cc-by-4.0
task_categories:
- object-detection
size_categories:
- n<1K
---
# Wgrape Unipd Bunch Detection

wGrapeUNIPD-DL contains lateral-view images of white grape varieties in vertical-shoot-positioned vineyards at six Italian locations and several phenological (BBCH) stages, taken 1.5-3 m from the canopy, with every grape bunch boxed in YOLO format. The Zenodo record holds 343 of them (the article reports 373): 241 size- and colour-calibrated with a Macbeth chart and 102 uncalibrated images taken before bunches formed, which have no boxes; calibration status, cultivar and BBCH stage from the folder names are kept as columns. The dataset contains 343 images with 1,955 bounding box annotations across 1 category.

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

## Citation

```bibtex
@article{sozzi2022wgrapeunipd,
  title={wGrapeUNIPD-DL: An open dataset for white grape bunch detection},
  author={Sozzi, Marco and Cantalamessa, Silvia and Cogato, Alessia and Kayad, Ahmed and Marinello, Francesco},
  journal={Data in Brief},
  volume={43},
  pages={108466},
  year={2022},
  publisher={Elsevier}
}
```


The dataset itself can be cited as:

Sozzi, M., Cantalamessa, S., Cogato, A., Kayad, A., &amp; Marinello, F. (2022). *wGrapeUNIPD-DL: an open dataset for white grape bunch detection* [Dataset]. Zenodo. https://doi.org/10.5281/ZENODO.6757555


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