Since ODM does not support calibration panels I’m wondering if anyone has a general workflow for preprocessing images with the calibration panel data before using ODM.
It seems like this is a consistent question that gets directed to a vague “you need to preprocess images for calibration with reflectance panels” workflow with no concrete steps. Sometimes specific products like Micasense or Mapir will have software/code for doing this. But that works for their products only, I’m looking for a generalized workflow that would work for any combination of cameras and calibration panels (assuming their ground-truth reflectance values were known).
If anyone had a workflow to share with me I would be happy to:
A) Repost it under the “tutorial” category
B) Add it to the documentation
This seems like something that could greatly benefit the community.
Even if you only have technical readings or references for me to look into I could make a simple python script for the community to use in future. Personally I have 2 points of confusion in general:
- What is the actual math that needs to be applied to every pixel in every image to calibrate multispectral images from a calibration panel?
- How do you ensure that the output of the calibration preprocessing can be used in conjunction with data from the DLS that ODM can use. I assume this isn’t a problem, but wouldn’t want to preprocess imagery in a way that makes it so ODM can’t use the DLS data anymore.
I’ve seen stuff like: “Surface Reflectance = slope*Digital Number - intercept” (I forget where) but do not completely understand how to convert that equation (or something like it) into computational action for preprocessing images. Is that like applying a linear regression of DNs from established Reflectance values of different sections of a calibration panel to every other pixel of every other image?
P.S. If someone has already covered this and I just couldn’t find it let me know and I’ll remove my post