Real-Time Apple Disease Detection and Classification Using Hybrid CNN Model

Abstract

Identifying and categorizing diseases in apple fruit is a difficult and time-consuming task in the field of agriculture. It is crucial to have an automated method for detecting apple diseases to effectively monitor and ensure sufficient and healthy production. While disease symptoms are visible in the apple fruit, having experts diagnose them in a lab is expensive and time consuming. This paper proposes a deep learning approach to detect and classify three types of common fungal diseases in apples (apple scab, apple rot, and apple blotch) from Red Green Blue (RGB) images of apples taken at various resolutions. The convolutional neural network model is used to distinguish between healthy and diseased apples. Agriculture heavily relies on digital image processing and analysis to ensure the production of high-quality fruits. Using CNN as a classifier to automatically detect and classify apple diseases, we have experimentally proven the importance of pre-programmed knowledge in the agriculture industry. Crossvalidation and testing on unseen data were conducted to exhaustively evaluate the trained model in various parameters. The experimental results have demonstrated that the proposed deep learning-based algorithm can accurately classify the three types of apple diseases with good accuracy.