Fruit Disease Detection using AI: A Review of Classical and Deep Learning Approaches

Fruit production is significantly impacted by diseases and manually inspecting crops at scale is not a feasible solution. In recent years, numerous studies have explored advanced computational methods for the identification and classification of fruit diseases using machine learning and deep learning technologies. Initially, simpler algorithms such as support vector machines (SVM) and artificial neural networks (ANN) showed promising results, yet they encountered limitations in feature extraction and generalization. However, the accuracy of these models notably improved with the introduction of advanced techniques like convolutional neural networks (CNNs), reaching up to 98.8% in real-time detection models for mango fruit diseases. This review highlights the depth of deep learning applications in addressing challenges like dataset diversity and model scalability, both of which are essential for advancing agricultural technologies. Serving as a rich knowledge base, this survey offers researchers and scholars a broad overview of both traditional and cutting-edge methodologies in fruit disease diagnosis. It can be regarded as a valuable academic resource for those engaged in agricultural AI research. Furthermore, it lays out a comprehensive foundation for future research and innovation in disease control within the agricultural sector by synthesizing existing strategies and identifying key areas for enhancement.