With the increasing global water scarcity, efficient assessment methods for crop water stress have become a prerequisite to perform precision irrigation scheduling. The 1accessibility of infrared thermal sensor provides a powerful tool to detect and quantify crop water stress. This paper reviews the current practices of infrared thermal imagery utilized to assess crop water stress. Overall, three technological aspects of infrared thermal sensing applications for crop water stress assessment are reviewed along with the challenges and recommendations: (i) introduction of uncooled thermal camera and platforms, including ground-based platform and unmanned aerial vehicles (UAVs) platforms, for thermal imaging acquisition, (ii) strategies of canopy segmentation in thermal imaging used to obtain average canopy temperature for CWSI calculation, (iii) correlation between three forms of crop water stress index (CWSI) i.e. theoretical CWSI (CWSIt), empirical CWSI (CWSIe), and statistic CWSI (CWSIs) and physiological indicators. The emphasis is on imaging process techniques for canopy segmentation in thermal imaging. As a future perspective, the potential use of deep learning approaches to assess crop water stress has been elaborated highlighting the future trends.

随着全球缺水量的增加，有效的作物水分胁迫评估方法已成为进行精确灌溉计划的前提。1红外热传感器的可及性为检测和量化作物水分胁迫提供了强大的工具。本文回顾了用于评估作物水分胁迫的红外热成像的当前实践。总体上，对用于作物水分胁迫评估的红外热传感应用的三个技术方面以及挑战和建议进行了回顾：（i）引入了非制冷热像仪和平台，包括地面平台和无人机平台，热成像采集；（ii）用于获得CWSI计算的平均冠层温度的热成像中的冠层分割策略，（iii）三种形式的作物水分胁迫指数（CWSI）之间的相关性，即理论CWSI（CWSSI），经验CWSI（CWSIe），统计CWSI（CWSIs）和生理指标。重点是在热成像中用于冠层分割的成像处理技术。作为未来的观点，已经详细阐述了使用深度学习方法评估作物水分胁迫的潜力，突出了未来的趋势。