micasense / imageprocessing

MicaSense RedEdge and Altum image processing tutorials
https://www.micasense.com
MIT License
257 stars 152 forks source link

Alignment of Bands in RedEdge-P #191

Closed MarioCruz22 closed 7 months ago

MarioCruz22 commented 2 years ago

Getting in trouble with metadata differences like Vignette and Radiance and so on... Anyone knows how to fit RedEdge-P into the tutorial available at https://micasense.github.io/imageprocessing/Alignment.html? Or any other tutorial that can teach me how to treat RedEdge-P data and align its bands?

RedEdge Vignetting Center : 639.29433505013424, 480.39791730098648 Vignetting Polynomial : -0.00035899158967688416, 3.8849091850333786e-06, -2.057088751909051e-08, 5.0152576116649375e-11, -5.827880227440714e-14, 2.5353631877162346e-17

RedEdge-P Vignetting Polynomial 2D Name : 0,0,0,1,0,2,0,3,0,4,0,5,1,0,1,1,1,2,1,3,1,4,2,0,2,1,2,2,2,3,3,0,3,1,3,2,4,0,4,1,5,0 Vignetting Polynomial 2D : 0.467640,0.501446,-0.288309,-0.448964,0.306178,-0.083137,1.384667,0.284610,-0.472609,0.215251,-0.015028,-0.951077,-0.142211,0.523393,-0.179298,-0.884722,-0.275812,-0.063004,0.511571,0.124509,-0.058204

MarioCruz22 commented 2 years ago

I was wondering if are there any support that helps me to create a vignetting correction matrix with the metadata provided by RedEdge-P sensors!

Vignetting Polynomial 2D Name : 0,0,0,1,0,2,0,3,0,4,0,5,1,0,1,1,1,2,1,3,1,4,2,0,2,1,2,2,2,3,3,0,3,1,3,2,4,0,4,1,5,0 Vignetting Polynomial 2D : 0.467640,0.501446,-0.288309,-0.448964,0.306178,-0.083137,1.384667,0.284610,-0.472609,0.215251,-0.015028,-0.951077,-0.142211,0.523393,-0.179298,-0.884722,-0.275812,-0.063004,0.511571,0.124509,-0.058204

andrematte commented 1 year ago

Hi @MarioCruz22 , I'm facing the same issues when trying to process images taken with a RedEdge-P using open-source libraries. Did you happen to find any solution for this?

MarioCruz22 commented 1 year ago

Yup! I did find a solution but I do not recall where.. u can take a loot to "https://forum.developer.parrot.com/t/vignetting-correction-sample-code/5614", as it seems close to what I have, however my results didn't seem quite well nor accurate with corrections. After all I just dropped vignetting corrections, for now.

MarioCruz22 commented 1 year ago

Not close to my laptop right now, but as soon as I get my code I'll post it here.

andrematte commented 1 year ago

Thank you very much, that would be really helpful!

HazelSCUT commented 1 year ago

Hi, I have also encountered such a problem. The data cannot be compatible with Rededge-P. How could you solve such a problem? It seems that it is impossible to open the Rededge-P images, not to mention later processing (e.g., alignment or batch processing): image image Looking forward to your reply. Many thanks.

alherca73 commented 11 months ago

Hello @MarioCruz22 , did you modify your scripts so that imagery from RedEdge-P or Altum-PT can be processed from DN to Reflectance values?

MarioCruz22 commented 9 months ago

I did not yet… Did you? I am now back to work with this library and will try to fix these incompatibilities. Have some or any of you already managed to get reflectance values from DN or the 'vignette_polynomial_2D_name' from 'Vignetting Polynomial'? I also noticed that one link that I have commented above is now broke. Sorry for that!