Precision Agriculture in Sugarcane: Review of Disease Detection and Severity Classification Methods

Significant yield losses have been experienced by sugarcane, due to the prevalence of leaf diseases, making the accurate detection and severity classification of these diseases essential for effective crop management and precision agriculture. In this comprehensive review, recent advancements in automated leaf disease detection and severity classification using machine learning and deep learning have been investigated. Research papers have been analysed based on disease types, methodologies used, datasets, and evaluation metrics. It is revealed by this study that deep learning models consistently outperform traditional methods in terms of accuracy and efficiency for both disease detection and severity classification. Among all the deep learning models evaluated, the highest accuracy is demonstrated by InceptionV4. Furthermore, promising results have been produced by the combination of machine learning, and deep learning techniques. However, significant challenges have been observed, including the limited availability of high-quality annotated datasets, regional variability in disease symptoms, insufficient adoption of hybrid and transfer learning models, underutilized multimodal inputs, and limited focus on severity classification. Addressing the identified challenges through improved datasets, advanced hybrid models, region-specific customization, and integration of multimodal approaches is considered essential for achieving more robust and scalable solutions. Precision agriculture can be significantly enhanced by the adoption of advanced techniques in disease monitoring, ensuring improved disease management, and promoting sustainable sugarcane cultivation.