"PlantVillage Plant Disease Classification Dataset"

"The dataset includes a massive dataset of plant disease-related leaf images, which are to be detected and classified, also known as the PlantVillage Leaf Disease Dataset. It includes more than 54,000 RGB pictures of healthy and diseased plant leaves in a high quality to reduce background noise and lighting diversity to capture images with controlled imaging conditions.The dataset consists of 14 crop species and 38 different classes, representing a particular plant-disease association or a healthy state. Some of the diseases that are represented in the dataset are bacterial, fungal and viral infections that are prevalent among agricultural crops. Expert annotation was used to provide image labels using visible disease symptoms, which provided ground truth reliability in supervised learning tasks.All photos have been represented by RGB and normalised to match the deep learning processes. The dataset does not implement artificial augmentation, and researchers can add task-specific preprocessing or augmentation approaches to the needs. The dataset has a clean structure, balanced labeling, and has been widely applied in the literature, thus it is a benchmark resource in training, evaluation, and comparison of computer vision models in recognizing plant diseases.PlantVillage offers an open-source research and educational dataset to assist research and teaching in the fields of precision agriculture, plant pathology, and vision-based crop health sensing."