---
dataset_info:
- config_name: augmented
  features:
  - name: image
    dtype: image
  - name: mask
    dtype: image
  - name: num_instances
    dtype: int32
  - name: source_image
    dtype: string
  splits:
  - name: train
    num_bytes: 930394414
    num_examples: 6000
  download_size: 930548701
  dataset_size: 930394414
- config_name: raw
  features:
  - name: image
    dtype: image
  - name: mask
    dtype: image
  - name: num_instances
    dtype: int32
  - name: split
    dtype: string
  splits:
  - name: train
    num_bytes: 830967257
    num_examples: 1500
  download_size: 831044478
  dataset_size: 830967257
configs:
- config_name: augmented
  data_files:
  - split: train
    path: augmented/train-*
- config_name: raw
  data_files:
  - split: train
    path: raw/train-*
  default: true
license: cc-by-4.0
task_categories:
- image-segmentation
size_categories:
- 1K<n<10K
---
# Occpepseg Yolo Occluded Segmentation

This dataset provides RGB images of pepper plants in field environments, specifically focusing on the challenge of segmenting occluded peppers within natural agricultural settings. Captured using an ORBBEC Gemini 335L camera during June-July 2025 in Inner Mongolia, China, the images depict real-world field conditions with varying degrees of foliage occlusion affecting pepper visibility. The dataset contains raw and augmented versions.
The raw split contains 1,500 images with pixel-level mask annotations.  

The augmented split contains 6,000 images with pixel-level mask annotations.  


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

```bibtex
@article{yu2026occpepseg,
  title={OccPepSeg-YOLO for Instance Segmentation of Occluded Peppers in Field Images},
  author={Yu, Xinran and Jiang, Mingxi and Gao, Fei and Fan, Yize and Bai, Yanyan and Peng, Zhigang and Lu, Qi and Liu, Qian and Xu, Shengyong},
  journal={Agronomy},
  volume={16},
  pages={1646},
  year={2026},
  publisher={MDPI}
}
```


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