This is the MSU Apple Dataset Version 2 repository. All images were taken in the orchard at the Michigan State University. This dataset consists of 1246 images including 'Gala' and 'Blondee' two varieties and also is split into 934 training sets and 312 test sets. All of data is organized in COCO object detection format. The dataset can be downloaded from https://zenodo.org/record/7925397#.ZF0BVC_MJ9c.

If you would like to use our data in your publication, please cite this dataset.

@article{chu2023o2rnet,
  title={O2RNet: Occluder-Occludee Relational Network for Robust Apple Detection in Clustered Orchard Environments},
  author={Chu, Pengyu and Li, Zhaojian and Zhang, Kaixiang and Chen, Dong and Lammers, Kyle and Lu, Renfu},
  journal={arXiv preprint arXiv:2303.04884},
  year={2023}
}