Advanced AI for Helianthus Disease Detection: A Hybrid CNN-KNN Model

The precise and prompt identification of diseases in sunflower crops is of utmost importance for the long-term viability of agriculture and the assurance of food supply. This study introduces an innovative methodology for the categorization of several diseases affecting helianthus plants, such as Rust, Wilt, Rot, and Stem Canker. The proposed strategy utilizes a hybrid deep learning model that integrates convolutional neural networks (CNN) and k-nearest neighbor (KNN) networks. By utilizing a varied dataset consisting of 25,000 high-resolution photos, our hybrid CNNKNN model attains a noteworthy overall accuracy rate of 95.78%. The incorporation of KNN layers in the model enables the consideration of temporal dynamics in the evolution of diseases, leading to a notable enhancement in the accuracy of disease classification. This improvement is particularly evident when differentiating between visually identical diseases, such as Rust and Stem Canker. The hybrid technique we propose has clear advantages when compared to solo CNN and KNN models, as evidenced by our comparative research. The research conducted in this study holds significant practical consequences as it presents a robust method for the automated diagnosis of diseases in sunflower crops. This tool has the potential to improve crop health and contribute to the promotion of sustainable agricultural practices.