Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix start/end date on Windows and enforce band order normalization #1701

Merged
merged 5 commits into from
Oct 2, 2023

Conversation

pierotofy
Copy link
Member

Fixes #1699

@pierotofy pierotofy changed the title Fix start/end date on Windows Fix start/end date on Windows and enforce band order normalization Sep 29, 2023
@pierotofy
Copy link
Member Author

This PR also enforces band order normalization, relying on the band_index provided by manufacturers provides inconsistent outputs, e.g. the Mavic 3M puts the RGB band at index 5, MicaSense at 1, plus countless other permutations found in the wild. It's a pain for users to then sort out why "RedEdge" is the first channel band when opened in QGIS (or worse, assume that ODM doesn't work because RGB is not the default band).

We now enforce this band order:

  • RGB
  • NIR
  • RedEdge
  • LWIR

Which is pretty common/standard.

@Saijin-Naib
Copy link
Contributor

Perfect. Follows the wavelength order too.

@pierotofy
Copy link
Member Author

pierotofy commented Sep 29, 2023

This will also improve multispectral alignment, because in cases where the manufacturer chose "RedEdge" as index 0, RedEdge was automatically chosen as the primary band (unless overriden via --primary-band, which few people use), giving poor alignment results.

We now automatically favor RGB, or a Blue band if available, otherwise the first band in the list.

@Saijin-Naib
Copy link
Contributor

Can we favor green over blue for alignment? That should be higher SNR than blue on these sensors.

@pierotofy
Copy link
Member Author

Sure!

@Saijin-Naib
Copy link
Contributor

Saijin-Naib commented Sep 29, 2023

I don't know how to modify to test, but most sensors mimic human vision and green is the most responsive channel, plus slightly longer wavelength gives less noise on CCD and CMOS (blue is too close to UV), but not so long focusing and scatter becomes problematic (NIR/red-edge)

So, it should be slightly better but I can't prove it out 😑

@pierotofy pierotofy merged commit f9a3c5e into OpenDroneMap:master Oct 2, 2023
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Fix capture start/end date on Windows
2 participants