Integrative Framework for Accurate Corriander Disease Classification: Unifying Deep Learning and Machine Learning Approaches with Feature-Rich Hybrid Models

The purpose of this study is to present a revolutionary methodology for the precise categorization of plant diseases. The methodology makes use of a combination method that combines advanced machine learning and deep learning approaches. For the purpose of accomplishing robust feature extraction, the model that has been developed incorporates resulting radial distances, features based on leaf perimeters, and RGB color moments. A Random Forest classifier is developed during the training phase by making use of these features, and a Convolutional Neural Network (CNN) is utilized as a feature extractor during this phase. The model is evaluated with the help of a dataset that contains photos of a variety of plant diseases. These diseases include powdery mildew, leaf spot, downy mildew, Cercospora leaf blight, bacterial leaf spot, Fusarium wilt, and Aster yellows. The remarkable performance of the model over a wide range of illness categories is highlighted by the thorough evaluation measures, which include precision, recall, and Fl-Score. It is noteworthy that the methodology that has been described obtains an amazing overall accuracy of 98%. The study makes use of a comprehensive confusion matrix and delivers metrics that are specific to each class. This provides a comprehensive insight of the algorithm's strengths as well as areas that could want significant improvement. Additionally, the research evaluates the proposed hybrid model in comparison to the most advanced leaf recognition techniques currently available, hence establishing the hybrid model's competitive advantage and its applicability for practical applications. The findings provide evidence that the model is reliable in the identification of plant diseases, highlighting the way in which it has the potential to be an invaluable instrument for precise farming and efficient crop management. In general, the integrated strategy that has been proposed makes a significant contribution to the field of plant disease classification by providing a solution that is dependable and accurate for the early diagnosis and intervention of diseases.