JujubeBruiseNet: A high-resolution image dataset for bruise detection in Ziziphus mauritiana.

The article presents JujubeBruiseNet, a high-resolution image dataset designed for bruise detection in Ziziphus mauritiana (jujube) fruits. Ziziphus mauritiana is a seasonal fruit often found in late summer to early fall. The bruise detection in this fruit is crucial for post-harvesting, fruit processing, and food packaging. Manual detection of bruises is time-consuming and often leads to inaccuracy. Therefore, developing a novel classification model is essential, which will immediately recognize bruises in the fruits and, as a result, decrease human effort, expenses, and production time in the agriculture sector. The dataset contains a total of 1464 original photos categorized by two classes labelled Healthy and Bruised. We collected the fruit from the local market and fields near Savar, Dhaka, Bangladesh, with the help of domain experts in the period from 10th March to 20th March 2025. To reduce outside variations and provide uniformity, the photos were taken under precisely controlled lighting. This article offers a major dataset for researchers to develop effective quality assessment models for post-harvesting fruit sorting and classification. Convolutional neural networks (CNNs) and other computer vision models can be trained exclusively using this dataset to increase the precision of agricultural product bruise recognition. The dataset can facilitate research in computer vision-based agricultural monitoring and fruit quality evaluation, openly accessible on Mendeley Data, link: JujubeBruiseNet: A Dataset for Bruise Detection in Ziziphus mauritiana - Mendeley Data.