<![CDATA[This study aims to develop a banana ripeness classification system using a Convolutional Neural Network (CNN) method based on Transfer Learning with MobileNetV2 and EfficientNetB0 architectures. The dataset used consisted of 1,142 banana images divided into training and validation data. The training process applied data augmentation and regularization techniques to improve the model generalization capability. The results showed that the MobileNetV2 model achieved the best performance with an accuracy of 97.75%, precision of 97.84%, recall of 97.75%, and F1-score of 97.74%, while EfficientNetB0 achieved an accuracy of 90.09%. The best model was implemented into a website-based prediction system for automatic and real-time banana ripeness classification. The results indicate that CNN based on Transfer Learning provides excellent performance in identifying banana ripeness levels.]]>
Copyrights © 2026
