Apple Leaf Disease Detection Using Deep Learning: A Hybrid Classification and Segmentation Approach

Apple leaf diseases create substantial obstacles for apple cultivation by reducing both total output numbers and the quality standards of final products. The current disease detection systems are based on manual inspections that produce inconsistent results due to human errors while creating substantial operation expenses. The proposed method applies deep learning architectures by merging U-Net with Convolutional Neural Networks to achieve automatic and precise apple leaf disease detection. The system initiates by determining image classification for healthy or diseased leaves before it proceeds to segment the afflicted areas. A training procedure enhancement leads to 97% correct classification outcomes and an 86% segmentation precision in the model. Through an online upload framework users submit leaf pictures so the system displays diagnostic results as well as marked affected areas. The system implements precision farming methodology by providing farmers a quick and reliable technology to discover early-stage diseases that strengthens crop well-being and production levels.