You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to COLMAP
Dear Johannes,
I was trying Colmap 3.0 on Gopro video with wide fov option. It appears from the code that the omega parameter should be estimated. Yet, the software always output 0 for omega. Can you please suggest if there are some silly mistake that I could have made? All the images were not intrinsically calibrated and I enabled shared intrinsic for all of them.
Thanks,
Johannes Schönberger
unread,
Aug 2, 2017, 5:09:52 AM8/2/17
Reply to author
Sign in to reply to author
Forward
Sign in to forward
Delete
You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to COLMAP
I see. I was setting it close to 0 and there was a threshold that you set in the distortion function that essentially outputs 0 gradient. My bad!
What is a reasonable initialization/procedure to deal with FOV camera in your experience because the nonsense initial guess will affect the correspondence estimation step in 2 view matching.
Johannes Schönberger
unread,
Aug 2, 2017, 5:20:17 AM8/2/17
Reply to author
Sign in to reply to author
Forward
Sign in to forward
Delete
You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to col...@googlegroups.com
Setting it omega=1e-3 should produce minimal distortion effects and hence not affect the two-view matching too much. If you want to be more precise, you could first set omega=0 during the matching and then omega=1e-3 before starting the reconstruction. Makes sense?