Transfer Learning Based Apple Fruit Disease Cllasification Using Pre Trained Moldes (Resnet, Efficientnet, Vgg, Mobinet)

Accurate and early detection of fruit diseases is essential for improving agricultural productivity and reducing economic losses. This research presents a transfer learning-based approach for apple fruit disease classification using state-of-the-art pre-trained deep learning models including ResNet, EfficientNet, VGG, and MobileNet. The proposed system leverages pre-trained convolutional neural networks to extract deep features from apple images and classify them into multiple disease categories. The experimental results demonstrate that transfer learning significantly improves classification accuracy while reducing training time and computational cost. Among the evaluated models, EfficientNet achieved the highest accuracy, followed by ResNet, MobileNet, and VGG. The proposed approach is suitable for real-time agricultural applications and can be deployed on mobile and edge devices. Agriculture remains the backbone of many economies, particularly in developing countries like India, where a significant portion of the population depends on farming for livelihood. Among horticultural crops, apple production holds substantial economic importance. However, apple cultivation is highly vulnerable to a variety of diseases such as apple scab (Venturia inaequalis), black rot (Botryosphaeria obtusa), and cedar apple rust (Gymnosporangium juniperi-virginianae). These diseases not only reduce yield but also degrade fruit quality, leading to major financial losses.