Demonstration of a Highly Developed CNN-SVM Model to Accurately Assess the Degree of Brown Rot in Orange Leaf

Through the use of a combination Convolutional neural network, also known as CNN, and a Support Vector Machine (SVM) model, this investigation tackles the significant problem of Brown Rot intensity in the leaves of orange. A detailed assessment of the model's effectiveness over a variety of severity levels is presented in this paper. The assessment makes use of precision, recall, F1-score, and overall accuracy measures. Based on the findings, it seems that the model is successful in properly assessing the degree of brown rot infection in orange leaves. In terms of severity level 1, the model obtains a precision of 95%, a recall of 97%, and an F1-score of 96%, resulting in an overall accuracy of 97%. A precision of 94%, a recall of 96%, and an F1-score of 95% are all shown by severity level 2 in terms of accuracy. A reliability rating of 97%, a recall of 93%, and an F1-score of 94.5% are all shown by the model when it is applied to severity level 3. An accuracy of 96%, a recall of 98%, and an amazing F1-score of 97% are all shown by the severity level 4, which is the most severe level. It has been observed that the overall accuracy across all severity levels amounts to 97%. These results provide further evidence that the hybrid CNN-SVM model is effective in accurately classifying the severity of Brown Rot, hence emphasizing the potential of the method for usage in agriculture precision and making crops lifehealthy disease control.