Beyond Black Boxes: An Ontology Based Vision Framework for Transparent Fruit Disease Detection

Efficiency and reliability in the timely identification of phytopathogens would help minimize crop loss in agricultural sectors but existing deep learning based methods, although accurate, would require a large annotated dataset and are black-boxed in their operation, thus, restricting flexibility a