Pomegranate disease detection and classification dataset for deep learning applications: A case study from Halabja city

Timely and accurate detection of pomegranate fruit diseases is critical for minimizing crop losses, preserving fruit quality, and supporting sustainable agricultural practices. This study introduces the Halabja Pomegranate Fruit Disease Image Dataset, a systematically compiled collection of images from orchards in one of Iraq's major pomegranate-producing regions. The dataset comprises 2178 original images and 28,314 augmented images, categorized into four specific classes: ectomyelois ceratoniae, colletotrichum spp., sunburn, and healthy fruit samples. To create an ecological setting and ensure significant class variation, images were captured in natural outdoor environments. A standard preprocessing step was applied, which involved resizing all images to 512×512 pixels and using several image augmentation techniques to improve the flexibility and robustness of machine learning models. The unique characteristics of this dataset make it highly suitable for developing machine learning and deep learning models aimed at plant disease detection and other computer vision tasks in precision agriculture. Its contextual relevance and content diversity make it valuable for building an effective diagnostic tool capable of functioning in real field conditions.