Suitability of sat-bundleadjust on Pleiades/PNeo images

[kaaviyave]

Hi team,

Thanks a lot for sharing your work and the paper!

I am working with multiple stereo-pairs from Pleiades and PNeo covering large areas in the order of 60 to 250+ sq kms in most cases.

For my preliminary tests, I am using 4 Pleiades images covering 190 sq kms (2 stereo-pairs acquired on 2 different dates).
I tried performing bundle adjustment using your packages (where I have tested so far the ba_bruteforce and ba_sequential modes).
The ba_bruteforce results in strange results whereas the ba_sequential mode succeeds to reduce the z-shift from 15m to 6m but the shift still remains quite significant.

I would like to understand if this could be due to my large study area and if the assumptions of single camera center would still hold in my use-case? If it wouldn't hold, would you be able to suggest some potential solutions/adaptations I could implement to improve the adjustment?

Thank you again for your time.