Disease Recognition in ViticultureCNN-SVM Framework for Accurate Grapevine Disease Diagnosis

This research paper introduces a novel methodology that employs Convolutional Neural Networks (CNNs) and Support Vector Machines (SVMs) to classify grapevine leaf diseases. A complete dataset consisting of photos of grapevine leaves exhibiting different illnesses and healthy conditions was gathered and subjected to preprocessing techniques to assure consistency and improve the resilience of the model. The design of the Convolutional Neural Network (CNN) comprised of three convolutional layers, succeeded by three max-pooling layers, and culminated in a solitary fully connected layer. The SVM classifiers were trained for disease categorization using extracted characteristics. The experimental findings provided empirical evidence supporting the efficacy of the suggested methodology, as it achieved a classification accuracy over 95% when applied to several grapevine diseases. This outcome highlights the potential of the strategy for accurate and reliable disease diagnosis in the field of viticulture.