Integrating Convolutional Neural Networks with Support Vector Machines for Enhanced Classification of Cabbage Leaf Diseases

Many diseases plague cabbage production, including Alternaria leaf spot, which causes production losses. When the ambient temperature is between 68 and 81°F and there is wetness on the cabbage, Alternaria is more prone to spread illness. Yellow dots that enlarge and form rings around them, like a target or bull's eye, are among the symptoms. There may be holes in the leaves as a consequence of the centers falling out when the tissue dies. The hybrid model proposed in this paper combines CNN and SVM techniques to reliably diagnose cabbage leaf diseases, with a focus on Alternaria leaf spot. The study used 1200 cabbage leaf photographs divided into training, validation, and testing sets. The hybrid model outperforms at 93.30% accuracy. When compared to CNN, SVM, and KNN models, the hybrid approach performs well. Performance metrics like accuracy, recall, and F1-score validate the model's robustness and reliability. The findings indicate that deep learning and traditional machine learning may be used to reliably identify cabbage illnesses, hence enhancing food security and agricultural sustainability.