Unmanned aerial systems (UAS) for collecting multispectral imagery of agricultural fields are becoming more affordable and accessible. However, there is need to validate calibration of sensors on these systems when using them for quantitative analyses such as evapotranspiration, and other modeling for agricultural applications. The results of laboratory testing of a MicaSense (Seattle, WA, USA) RedEdge™ 3 multispectral camera and MicaSense Downwelling Light Sensor (irradiance sensor) system using a calibrated integrating sphere were presented. Responses of the camera and irradiance sensor were linear over many light levels and became non-linear at light levels below expected real-world, field conditions. Simple linear corrections should suffice for most light conditions encountered during the growing season. Using an irradiance sensor or similar system may not properly account for light variability in cloudy or partly cloudy conditions as also identified by others. A simple stand for aiding in reference panel imaging was also described, which may facilitate repetitive, consistent reference panel imaging.

中文翻译：

---

用于定量农业应用的通用短波多光谱相机系统的校准

用于收集农田多光谱图像的无人机系统 (UAS) 变得越来越便宜和容易获得。但是，在将这些系统上的传感器用于定量分析（例如蒸发蒸腾）和其他农业应用建模时，需要验证这些系统上传感器的校准。展示了使用校准积分球对 MicaSense（西雅图，华盛顿州，美国）RedEdge™ 3 多光谱相机和 MicaSense 下井光传感器（辐照度传感器）系统进行的实验室测试结果。相机和辐照度传感器的响应在许多光照水平下是线性的，在低于预期的真实世界现场条件的光照水平下变得非线性。对于生长季节遇到的大多数光照条件，简单的线性校正就足够了。使用辐照度传感器或类似系统可能无法正确解释多云或部分多云条件下的光变化，正如其他人所识别的那样。还描述了一个用于帮助参考面板成像的简单支架，它可以促进重复、一致的参考面板成像。