Deep Learning for Plant Health: Accurate CMV Morbidity Scale Classification Using Faster R-CNN

Cucumber mosaic virus (CMV) is a newly discovered virus that has the potential to severely reduce crop yields worldwide. To effectively manage a condition, a timely and precise illness categorization based on the CMV morbidity scale is required. Using deep learning (DL) methods, the study presents a fresh strategy for the multi-classification of CMV-infected plant samples. The suggested study gathered 15,000 photos from a wide range of crops at various phases of CMV infection. The morbidity scale labels were thoroughly documented in the dataset, making it easier to train the DL model. With its superior performance in object identification tasks, the Faster Region-based Convolutional Neural Network (R-CNN) architecture was adopted for CMV classification. The total accuracy of the proposed DL model in categorizing CMV-infected plants according to their morbidity scale categories was an outstanding 98.11%. The superiority of the proposed method was demonstrated by comparison to baseline models such as conventional machine learning classifiers and object identification models. Farmers and plant pathologists may now make more well-informed decisions more quickly because of the Faster R-CNN model’s successful deployment, which provides a solid and automated method for diagnosing CMV illness. Classifying plants infected with cytomegalovirus accurately helps optimize crop protection techniques, which in turn improves global food security. The findings highlight the need for accurate and timely disease categorization for sustainable agriculture and point to the potential of DL algorithms to revolutionize plant disease diagnosis. Innovative methods may be used to successfully combat CMV and improve agricultural practices everywhere.