Hi; that's an awesome piece of work: thanks for posting.
I do also thermal orthophotos of terrains and other targets (such as Photovoltaic panels - attached a sample brochure showing a PV installation inspection; over 250 images were collected for the mosaic) and to my experience and knowledge it's hard to obtain good results, especially if radiometric data are expected (low sensor resolution, low image contrast, artifacts, temperature changes between images, radiometric alterations that might be introduced by the processing software, etc.).
I have some questions:
- What IR camera did you use? What's the resolution?
- Did you manually trigger the camera or was it automatically triggered by the flight controller?
- Were IR images georeferenced? If yes as I imagine, was this done in real time during acquisition, or in post?
- What software did you use to obtain the thernal ortho-photo?
- Is the ortho-photo radiometric?
- I noticed you captured 'R-G-B' palette'd images: what band-weight did you chose?