The WebODM documentation for processing multispectral drone imagery says to pass the –radiometric-calibration parameter. It does not say that any preprocessing of the images is necessary. However, a number of discussions on here state that some preprocessing of the images is necessary.
Can someone clarify what preprocessing is required to create orthomosaics from multispectral drone images? It would be helpful if this was added to the documentation as well.
No pre-processing is necessary, unless you want to radiometrically calibrate the images yourself (for example, if you have reflectance panels and want to use those).
If I do want to preprocess to calibrate based on reflectance panels should I not pass the radiometric-calibration parameter? What if I preprocess for the calibration panels but also want to process the DLS compensation?
If I want to preprocess for the calibration panels does it make sense to preprocess all the way to stacked images (thus --skip-band-alignment) or only do the minimal preprocessing (calibration panels)?