Deep Learning Approaches for Apple Leaf DiseaseDetection and Classification: A Comprehensive Review

Abstract

foliar diseases pose a major threat to apple cultivation, severely affecting crop yield and quality. Conventional disease detection relies on manual expert examination—a slow, subjective, and error-prone process. Recent advances in deep learning, especially Convolutional Neural Networks (CNNs), have demonstrated remarkable capability to automate plant disease detection with high precision. This review comprehensively surveys state-of-the-art deep learning models for apple leaf disease detection, covering research published between 2021 and 2025. We compare architectures such as Dense Net, Efficient Net, Mobile Net, Inception-V3, and Vision Transformers with respect to performance metrics, limitations, and real-world applicability. Current models achieve 90–98.32% classification accuracy, yet significant challenges remain in dataset diversity, field generalization, early-stage detection, and edge deployment. The review provides actionable directions for future research in precision agriculture.