APPLE LEAF DISEASE CLASSIFICATION USING DEEP CNNS

Abstract

The presence of foliar diseases in apple crops poses a significant threat to both yield and fruit quality, necessitating the development of reliable automated detection mechanisms. This study presents a deep learning–based framework employing a Convolutional Neural Network (CNN) for the accurate classification of apple leaf conditions into four distinct categories: Apple Scab, Black Rot, Cedar Apple Rust, and Healthy. A total of 3,171 annotated images were sourced from the PlantVillage dataset and divided into training (70%), validation (20%), and testing (10%) subsets. The proposed CNN model incorporates four convolutional modules, each consisting of batch normalization and max-pooling layers, followed by a fully connected dense layer with a dropout rate of 0.5 to mitigate overfitting. Model training was conducted for 40 epochs using the Adam optimization algorithm. Experimental evaluation demonstrated that the network attained an overall classification accuracy of 96.85% on the test dataset, with F1-scores ranging from 93.8% to 97.9% across the individual disease classes. The findings confirm that the proposed CNN architecture provides a robust and efficient approach for early and automated detection of apple leaf diseases, thereby supporting precision agriculture and timely crop management interventions.