TRANSFER LEARNING AND DEEP LEARNING METHODS TO PLANT DISEASE DETECTION: A SYSTEMATIC REVIEW

Plant diseases cause significant losses in agricultural productivity and threaten the worldwide food supply.This rapid development in the field of artificial intelligence, especially deep learning, has made Convolutional Neural Networks (CNNs) potent tools for autonomous identification of plant diseases.This review examines the development of deep learning architectures, transfer learning approaches, model optimization, interpretability, and deployment systems for plant disease detection.Special emphasis is placed on the DenseNet architecture because it improves classification performance, computational efficiency, and feature reuse.The paper provides guidelines for future research studies to develop robust, real-time plant disease detection systems.