Multi-Crop Leaf Disease Detection with LLM-Assisted Diagnosis and Explainable AI

Abstract

Early identification of plant diseases is crucial for maintaining agricultural productivity and ensuring food security. Manually inspecting crop leaves takes a lot of time and requires expertise, making automated disease detection systems very helpful. This paper presents a framework for detecting leaf diseases in multiple crops that combines deep learning-based classification, explainable artificial intelligence, and large language model-assisted diagnosis. The system uses MobileNetV2 as the main convolutional neural network to classify leaf diseases from three major crops: potato, maize (corn), and grapes. The dataset is split into 80 percent for training and 20 percent for validation and testing. To improve the transparency of the predictions, Grad-CAM highlights important areas of the leaf that affect the model's decisions. Additionally, a large language model module creates human-readable diagnostic explanations and treatment suggestions for the detected diseases. Experimental evaluation shows that the proposed system achieves high classification accuracy and enhances the interpretability of the model predictions. The combination of explainable AI and LLM-based recommendations offers a useful decision-support tool for agricultural applications.