A Comprehensive Survey on Intelligent Techniques for Early Paddy Disease Detection and Diagnosis
Paddy is one of the key staple crops which support the livelihood of billions of people throughout the world. However, the productivity of this crop is constantly under threat from fungal, bacterial, viral and nutritional associated illnesses which cause huge losses in production and economy. Thus, early and precise diagnosis of illness is important for effective crop management and sustained agricultural production. Recently, automatic detection of paddy diseases has been improved greatly with the rapid progress of artificial intelligence (AI), computer vision, deep learning, remote sensing, Internet of Things (IoT) and Unmanned Aerial Vehicle (UAV)-based monitoring. In this survey, the recent works in intelligent diagnosis of paddy diseases have been comprehensively studied by considering conventional diagnosis approaches, machine learning approaches and deep learning approaches including convolutional neural networks, vision transformers, hybrid models and explainable AI frameworks. The paper also summarises the publicly accessible datasets, picture pre-processing methods, feature extraction approaches, classification algorithms, assessment metrics and deployment platforms for field-level disease monitoring. The previous studies are compared for diagnostic performance, computational complexity, resilience, and practical usefulness under real-world agricultural situations. The survey also addresses the present issues such as scarce and imbalanced data, environmental variations, similarity of disease symptoms, model interpretability, generalisation across domains, and deployment with limited resources. Finally, we discuss current research themes like as multimodal data fusion, federated learning, foundation models, edge intelligence, digital twins and autonomous precision agriculture as interesting approaches for next generation paddy disease diagnosis systems. This assessment, by summarising recent developments and pointing out crucial research voids, is an important reference for researchers, agricultural engineers, and practitioners aiming to find accurate, scalable and intelligent solutions for crop disease management.