# Herbicide-Damaged Waterhemp (Amaranthus tuberculatus) Greenhouse Images: Healthy and 2,4-D, Glyphosate, Glufosinate and Fomesafen Injury Symptoms (Ohio State University)

## Facts

- Provider: Lalit Pun Magar, Matthew Thompson and Eugene P. Law (The Ohio State University)
- License: CC0-1.0 (https://huggingface.co/datasets/punlalit19/Waterhemp_GreenHouse_Images)
- Size: 624 files
- Format: JPEG (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/punlalit19/Waterhemp_GreenHouse_Images
- Access: public, readable without an account
- Category: Agriculture & Environment
- Source: https://huggingface.co/datasets/punlalit19/Waterhemp_GreenHouse_Images
- Page: https://scigantic.com/archive/9ce55192-53da-58c1-8548-83e49b9b6a9b
- Keywords: Agriculture & Environment, JPEG, Hugging Face

## About

Top-down photographs of potted waterhemp (Amaranthus tuberculatus), a herbicide-resistant weed of Ohio corn and soybean, grown in the Kottman Hall greenhouse at The Ohio State University and treated with 2,4-D (WSSA Group 4), glyphosate (Group 9), glufosinate (Group 10) or fomesafen (Group 14), with untreated controls. Collected by Lalit Pun Magar in the Weed Ecology and Management Lab to support computer vision rating of herbicide injury, which is otherwise scored visually by eye, and used at FloraPalooza 2026 to evaluate BioCLIP, BioCLIP 2 and DINOv3 image embeddings. There are 622 RGB JPEG images at 4032 x 3040 taken from 30 cm and 50 cm above the pot, sorted into folders by symptom class: healthy_untreated, chlorosis_only, necrosis_only, chlorosis_necrosis, epinasty and foliar_distortion. File names encode date, trial, camera height and pot id, so the same plant can be followed over time. About 4.3 GB; load as an image folder with datasets or torchvision. Useful for weed phenotyping, herbicide injury classification and plant stress detection. Dedicated to the public domain under CC0 1.0 by the curators; please cite the dataset.

## Load the data

```python
from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="punlalit19/Waterhemp_GreenHouse_Images", filename="chlorosis_necrosis/2026-05-11_IGPTRIAL1H30_110_1_rgb_32effab8-e969-4b6b-ae19-f9168d37996a.jpg", repo_type="dataset")
print(path)
```

## Data structure

File types in sampled files: JPEG

Example file: `chlorosis_necrosis/2026-05-11_IGPTRIAL1H30_110_1_rgb_32effab8-e969-4b6b-ae19-f9168d37996a.jpg`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/9ce55192-53da-58c1-8548-83e49b9b6a9b
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/9ce55192-53da-58c1-8548-83e49b9b6a9b
- Over MCP: `get_schema_card` with id `9ce55192-53da-58c1-8548-83e49b9b6a9b` at https://api.scigantic.com/mcp

## Related topics

- [JPEG datasets](https://scigantic.com/topics/jpeg)
- [Agriculture & Environment datasets](https://scigantic.com/topics/agriculture-and-environment)
