Hybrid Potato Tuber Dataset (Solanum tuberosum, Solanum phureja)
收藏Zenodo2026-06-27 更新2026-06-12 收录
下载链接：
https://zenodo.org/doi/10.5281/zenodo.20616991
下载链接
链接失效反馈官方服务：
资源简介：
The dataset comprises 36,000 images in .JPG format with a resolution of 224×224 pixels, organized into two classes: Good Condition and Defective.
The Good Condition class contains 18,000 images of potato tubers from the chaucha (Solanum phureja) and chola (Solanum tuberosum) varieties, all without visible superficial defects.
The Defective class also includes 18,000 images, equally distributed across both varieties (9,000 per variety). Each subset is further divided into three categories of superficial defects:
Manually induced surface cuts.
Sprouted tubers.
Rotten tubers.
The dataset is split into 64% for training (23,040 images), 16% for validation (5,760 images), and 20% for testing (7,200 images).
Additionally, several public datasets were incorporated to enrich this dataset, including:
Healthy Potato Image Dataset
Potato Disease Recognition Dataset
PotatoCare: Deep learning based potato disease dataset
提供机构：
Zenodo创建时间：
2026-06-09
