AI-powered Paralichthys olivaceus Disease Symptom Detection and Classification Method

In this study, we propose a deep learning-based disease symptom detection and machine learning-based disease classification technique to quickly diagnose Paralichthys olivaceus diseases. The disease symptom detection technique, using YOLOv9 model, showed excellent accuracy of about 85.4% of mAP and 75.2% of F1-score. The disease classification, an ensemble model using SGD classifier and random forest classifier was used, and the overall good correct answer rate was about 72% for Top-1, 83% for Top-2, and 94% for Top-3. As a result of testing by combining the two algorithms, when the threshold was not set in the disease classification result, it showed good performance of 96%. The Paralichthys olivaceus disease symptom detection and classification technique proposed in this study can take quick and effective disease treatment and prevention measures by the two algorithms working together to accurately detect disease symptoms and classify disease information.