Abstract
Accurate plant disease detection and crop recommendation are essential for sustainable agricultural practices. However, diagnosing plant diseases and determining suitable crops based on environmental conditions can be challenging and time-consuming. To address this, we propose an automated approach that integrates deep learning and machine learning techniques for plant disease classification and crop recommendation, with the detected disease and suggested crop further processed by an LLM (GPT) to provide treatment recommendations and crop insights. Specifically, we employ Swin Transformer, a state-of-the-art deep learning model, for precise detection of plant diseases from leaf images. Additionally, we use XGBoost to recommend crops based on topographical and soil structure data, utilizing a rich dataset from India's agricultural regions. In this paper, we explore the architecture and performance of our disease detection and crop recommendation models. Experiments on real-world datasets, including CCMT, demonstrate that our application provides accurate disease diagnostics and effective crop recommendations. Moreover, the system is deployed as a web application, enabling user-friendly access for farmers and agricultural professionals, and facilitating data-driven decision-making in plant health management and crop cultivation.