Agricultural water accounts for more than 70 % of the total global water usage, and the scarcity of global freshwater resources will largely limit global agricultural production. Precision irrigation is the key to improving water efficiency and achieving sustainable agriculture. Accurate and rapid access to crop water information is an essential prerequisite for precise irrigation decisions. Traditional moisture detection methods based on soil moisture and crop physiological parameters are faced with the problems of variable field conditions, low efficiency and lack of spatial information, which can be extremely limited in practical applications. By contrast, unmanned aerial vehicle (UAV) remote sensing has the advantages of low cost, small size, flexible data acquisition time, and easy acquisition of high-resolution image data. Therefore, UAV remote sensing has become an easy and efficient method for crop water information monitoring. This study systematically introduces the principles, methods and applications of crop water stress analysis using the UAV technology. First, the mechanism of crop water stress analysed by UAV is elaborated, focusing on the relationship between canopy temperature, evapotranspiration, sun-induced chlorophyll fluorescence (SIF) and crop water stress. Next, various UAV imaging technologies for crop water stress monitoring are presented, including optical sensing systems, red, green and blue (RGB) images, multi-spectral sensing systems, and hyper-spectral sensing systems. Subsequently, the application of machine learning algorithms in the field of UAV monitoring of crop water information is outlined, demonstrating their potential for data processing and analysis. Finally, new directions and challenges in UAV-based crop water information acquisition and processing are synthesised and discussed, with special emphasis on the prospects of data assimilation algorithms and non-stomatal restriction in monitoring crop water information in the future. This study provides a comprehensive comparison and assessment of the mechanisms, technologies and challenges of UAV-based crop water information monitoring, providing insights and references for researchers in related fields.

中文翻译：

---

基于无人机遥感系统的作物水分胁迫检测


农业用水占全球总用水量的 70% 以上，全球淡水资源的稀缺将在很大程度上限制全球农业生产。精准灌溉是提高用水效率和实现可持续农业的关键。准确、快速地获取作物水分信息是精确灌溉决策的重要前提。传统的基于土壤水分和作物生理参数的水分检测方法面临着田间条件多、效率低下、缺乏空间信息等问题，在实际应用中可能受到极大的限制。相比之下，无人机 （UAV） 遥感具有成本低、体积小、数据采集时间灵活、易于获取高分辨率图像数据等优点。因此，无人机遥感已成为一种简单高效的农作物水分信息监测方法。本文系统介绍了利用无人机技术进行作物水分胁迫分析的原理、方法和应用。首先，阐述了无人机分析作物水分胁迫的机理，重点介绍了冠层温度、蒸散、太阳诱导叶绿素荧光 （SIF） 与作物水分胁迫之间的关系。接下来，介绍了用于作物水分胁迫监测的各种无人机成像技术，包括光学传感系统、红、绿、蓝 （RGB） 图像、多光谱传感系统和高光谱传感系统。随后，概述了机器学习算法在无人机监测农作物水分信息领域的应用，展示了它们在数据处理和分析方面的潜力。 最后，综合并讨论了基于无人机的农作物水分信息获取和处理的新方向和挑战，特别强调了数据同化算法和非气孔限制在未来农作物水分信息监测中的前景。本研究对基于无人机的农作物水分信息监测的机理、技术和挑战进行了全面的比较和评估，为相关领域的研究人员提供了见解和参考。