Hybridizing CNN and SVM for Precise Arhul Flower Disease Diagnosis and Classification

This study uses a hybrid technique involving Convolutional Neural Networks with Support Vector Machines to construct a reliable illness detection model for Arhul flowers. Three layers of convolution, three max-pooling layers, or two fully connected layers make up the model architecture, which facilitates efficient feature extraction as well as learning from the image input. The main goal of this work is to develop a reliable and effective technique for foretelling different Arhul flower disorders. Metrics for recall, precision, F1-score, support, or accuracy are used to assess the model's performance. With values that range from 82.93 % to 88.64
^%
, the evaluation's findings show great precision across several illness groups. This demonstrates the model's ability to categorize occurrences of each disease appropriately. Recall values for each disease range from 83.12% to 87.06%, demonstrating the model's capacity to identify a sizable fraction of true positive instances. The F1 scores, which range between 84.47% to 87.15%, show an evenly distributed ability in terms of recall and precision. In light of the dataset's class distribution, the model's weighted average accuracy, which measures how well it predicts Arhul flower illnesses generally, is 85.50%. Even with unbalanced data, where some illness classes have more instances than others, the model performs consistently. The results of this study make a contribution to the field of farming disease control by providing a workable method for the early identification and classification of illnesses affecting Arhul flowers. In summary, this study offers a hybrid CNN-SVM approach to precise and effective illness prediction in Arhul flowers. The model performs well, demonstrating its promise for real-world agricultural applications that would ultimately enhance crop health tracking and oversight.