Prompt and accurate identification of crop diseases is essential for maintaining high crop yields and reducing economic losses. Traditional methods like PCR and pesticide application are effective but costly and time-consuming. This paper introduces AgriNet, a system that leverages MobileNetV2 architecture for efficient plant disease diagnosis. MobileNetV2’s advanced structure allows precise identification of leaf diseases, while explainable AI techniques such as SHAP ensure transparency in predictions. The integration of drone imaging enhances AgriNet’s applicability for real-time monitoring over large agricultural areas. Our contributions include the development of an accurate, interpretable, and scalable system for plant disease diagnosis, demonstrating its effectiveness on the PlantVillage dataset with a validation accuracy of 95%.