Panel captures and typical in-flight captures have different requirements in order to produce radiometrically calibratable data. In order to produce the best data possible, we offer two capture modes on our cameras - one designed for expected normal use and one designed specifically for capturing panel images. These modes adjust the gain and exposure time of each capture to get the proper exposure for the desired subject.
Panel Capture
For a panel capture, it is critically important that the panel is properly exposed. Overexposing a panel means that all of the pixels will appear white, but there is no way to tell how bright they really were - only that they were at least as bright as the maximum value for the given exposure and gain. MicaSense reflectance calibration panels are significantly more reflective than plants, so when taking a panel capture, using the manual capture button on the camera or in the WiFi interface or with the automatic panel detection feature, the camera will try and keep as much of the scene from saturating as possible, even if many pixels become underexposed. This makes it much more likely that a MicaSense reflectance calibration panel will be properly exposed. It is also possible to programmatically change the exposure settings for a panel capture by setting anti_sat=true with the HTTP API.
Here is an example of an overexposed panel image vs. a properly exposed panel image:
Normal Capture
For normal captures, such as when using an automatic triggering option, the goal is to have as many properly exposed pixels of the field as possible. To this end, the camera will try to have a large number of properly exposed pixels, potentially at the expense of some small areas that are overexposed or underexposed. This means that if a scene has a particularly bright spot, such as a specular reflection off of a car windshield or retaining pond, the camera will allow that spot to overexpose, so long as the majority of the image is still well exposed.
It is technically possible to get good panel shots with “normal capture” mode, and good field images with “panel capture” mode. As long as the captured images have their subjects properly exposed, the data will be useful. However, using the correct mode for the correct scene will generally lead to better data outputs with less hassle, which is why we recommend using the right capture type for the right scene.