ResNet-50 for Apple Leaf Disease Classification: Scalable and Accurate Automated Diagnostics

By influencing crop yield and quality, Apple diseases—including apple scab, black rot, and cedar apple rust—cause serious problems to world apple output. Conventional approaches of illness identification depend on subjective, labor-intensive, often error-prone manual inspection. This work provides a deep learning-based framework using the pre-trained ResNet-50 model to categorize apple leaf images into four categories: apple scab, black rot, cedar apple rust, and healthy. High-resolution images made up the dataset; it was methodically preprocessed utilizing resizing, normalizing, and data augmentation techniques including rotation, cropping, and flipping to improve model generalization and lower overfit. The model attained remarkable performance, with an accuracy of 98.54% on the test dataset, by substituting a bespoke dense layer and using transfer learning for the totally connected classification layer. Precision, recall, and F1-score among other metrics confirmed the model's general resilience across all classes. These results highlight how well deep learning models—such as ResNet-50— offer scalable, dependable, automated solutions for apple disease diagnosis, hence transforming agricultural diagnostics.