SMART DETECTION OF CROP DISEASES: UAV APPLICATIONS IN THE ERA OF PRECISION PLANT PATHOLOGY

Abstract

Against the backdrop of global food security concerns and the impending threat of phytopathogens, precision plant pathology technology has emerged as a key tool in ensuring the optimisation of agricultural sustainability. The conventional methods adopted in crop disease diagnosis, based on visual examination and laboratory analysis, are found to be lacking in terms of providing timely, geographically precise and scalable solutions. With recent advances in Unmanned Aerial Vehicle (UAV), or drone, technology, there is a paradigm shift in meeting such challenges. This review discusses in depth the synergistic integration of UAVs with multispectral, hyperspectral, thermal, and RGB imaging modalities in conjunction with artificial intelligence (AI) and deep learning approaches for the detection, classification, and quantification of diseases in plants at an early stage. Machine learning algorithms and optical sensors on unmanned aerial vehicles (UAVs) enable real-time high-resolution monitoring of disease signs on large crop fields. Vegetation indices, thermal stress maps, and spectral signatures are used by these systems to detect subtle physiological changes in crops before any visible sign of the disease. Their uses include disease detection, irrigation optimization, nutrient mapping, aerial sowing, yield prediction and precision pesticide application. Deep learning models, particularly CNNs and U-Net architectures, show the high accuracy of disease diagnosis and the estimation of their severity in field scenarios. In addition, UAV-based systems are fully compatible with Geographic Information Systems (GIS), IoTs, and cloud platforms, which facilitate data-driven decisionmaking for crop management. Nevertheless, there are obstacles in the shape of high data acquisition costs, model generalizability, regulatory restrictions, and low dataset diversity. The current article is concerned with recent developments, field-scale case studies, and existing challenges and discusses future directions for drone-based plant disease monitoring. The integration of UAV technologies with AI is highly promising to change the face of plant pathology and render disease monitoring more accurate, proactive, and sustainable.

Plant phenotyping relevance

UAV画像・分光/熱センシングとAIによる植物病徴の検出・重症度推定を中心に扱うレビューであり、植物状態の表現型取得手法が主題である。

Code and data availability

This is a review article on UAV-based crop disease detection. It presents no original phenotyping measurements, datasets, images, code, or models of its own; all cited works are prior publications, and no author-deposited public asset is mentioned.

No evidence-backed public reproduction asset is currently recorded.