For people interested about this work/dataset I recommand to check the Global Wheat Dataset : A more large and diverse dataset for wheat head detection http://www.global-wheat.com/ https://zenodo.org/record/5092309#.YaPEh9DMJPY The Wheat Ears Dection Dataset has been integrated in Global Wheat Dataset
Dataset from the Ear density estimation from high resolution RGB imagery using deep learning technique paper
[Simon Madec], [Frederic Baret], [Benoit de Solan], [Shouyang Liu]
The Wheat-Ears-Dection-Dataset (WEDD) is a dataset is a image dataset designed fo wheat ears detection in field condition.
- 236 high resolution images images (6000*4000)
- Wheat ears annotated with a bounding box
- 30729 ears identified
- Spatial resolution (GSD) of 0.13mm/pixel
- Two images for each microplots
- 20 contrasted genotype with 6 replicated growth in two environment
To cite the paper :
Madec, S., Jin, X., Lu, H., De Solan, B., Liu, S., Duyme, F., et al. (2019). Ear density estimation from high resolution RGB imagery using deep learning technique. Agric. For. Meteorol. 264, 225–234. doi:10.1016/j.agrformet.2018.10.013.
Dataset avalaible here
Please download replicate information along with images used for training and testing [here]
Below we present results.
| Method | Date | Source | AP | rRMSE | R² |
|---|---|---|---|---|---|
| Faster-RCNN [1] | 21/10/2018 | [2] | 0.85 | 5.3% | 0.91 |
The LabelIMG tool were used, please refer to this repository.