Sunflower Sentry: Advanced Hybrid Model for Early Disease Detection

Improving agricultural output and guaranteeing food security depends on early identification of plant diseases and precise categorisation of them. In this work, seven major diseases of sunflower such as Downy Mildew, Rust, Verticillium Wilt, Phoma Black Stem, Alternaria Leaf Spot, Sclerotinia Stem Rot, and Phomopsis Stem Canker are considered where From the advantages of both methodologies, the suggested model consists of Random Forest (RF) as well as Convolutional Neural Networks (CNN). Two sets of convolutional and max pooling layers make up the CNN architecture; then, a flattening layer. After that, a Random Forest with 14 trees—each with a maximum depth of 8—classified the obtained characteristics. From several web sources, 4,356 photos of sunflower leaves were gathered and preprocessed using cleaning, resizing, and augmentation. With an overall accuracy of 81.54%, the model was quite strong in identifying and grouping the disorders. For Phomopsis Stem Canker and Sclerotinia Stem Rot, accuracy is shown especially by precision, recall, and F1-score measurements. The confusion matrix study emphasised, even more, the strengths and opportunities for the development of the approach. Early intervention and efficient disease control made possible by this hybrid model provide a great tool for contemporary agriculture, hence allowing sustainable farming methods and maybe higher crop yields.