Exploring the Application of Drone Imagery in Early Identification of Crop Diseases and Pests

Abstract

Early and accurate identification of crop diseases and pests is critical to ensuring food security and sustainable agricultural development. The rapid advancement of high-resolution drone remote sensing technology provides innovative tools for early pest and disease detection. This paper explores the research progress, technical bottlenecks, and future directions of drone imagery technology in crop disease and pest monitoring. The study concludes that multi-spectral, thermal infrared and RGB sensors integrated on drone platforms can collaboratively capture centimeter-level high-resolution data. Through multi-source fusion of spectral, texture, and temporal data combined with lightweight model deployment, early spectral and morphological characteristics of crop stress caused by diseases and pests can be accurately identified, significantly improving detection accuracy compared to traditional satellite remote sensing and single machine learning methods. Case studies demonstrate that drone technology achieves 85%–95% recognition accuracy in monitoring typical diseases such as wheat rust and rice blast while reducing field inspection costs by over 60%. This paper provides a theoretical framework and technical roadmap for precision agriculture, offering practical significance for promoting agricultural digital transformation.

Code and data availability

This is a narrative review of drone-based crop disease detection with no original phenotyping data, imagery, code, models, or supplements. All cited datasets/works (e.g., AgriPest-1M, referenced DOIs) are prior work or hypothetical future proposals, not paper-specific public assets.

No evidence-backed public reproduction asset is currently recorded.