Bean Plant Pathologies Dataset for Deep Learning Tasks

This dataset is part of the Makerere University Beans Image Dataset, designed to diagnose bean crop diseases and conduct spatial analysis, available on this link (https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/TCKVEW). It includes images categorized into four classes: healthy bean leaves, Angular Leaf Spot (ALS) in bean leaves, Bean Rust in bean leaves, and an additional 'unknown' class. These images help in classifying and detecting between the target bean leaf classes and other visuals. The dataset was used for the project that leverages edge computing and deep learning for the real-time identification of bean plant pathologies. The dataset is organized into two main folders, each serving a specific purpose. The first folder contains data for the classification task, with images distributed among the four classes: healthy, ALS, Bean Rust, and the unknown class. The second folder is dedicated to the detection task, featuring annotations for ALS and Bean Rust, as well as unlabeled healthy images to enhance the learning of detection models.