Potato Knowledge Hub

A multi-view RGB image dataset of four potato varieties with visually assessed scab and dry-rot severity under dual-background imaging conditions.

Data in brief | 2026-Aug

This article presents a multi-view RGB image dataset of 9,106 potato tuber images across four varieties (Rumana, Asterix, Lal Pakri, and Diamant) captured under dual-background conditions. Tubers are categorized into 18 classes based on variety and visually assessed severity of common scab and dry-rot diseases. The dataset supports the development of computer vision algorithms for automated potato tuber phenotyping and post-harvest disease detection.