Question about correcting duplicate structure for SfM models

[zl...@cornell.edu]

Hi,

I found COLMAP usually is not able to correct duplicate structures of Internet landmarks photos. However, for my project, I have to have correct poses estimates for these models (or at least not have such incorrect poses estimate caused by duplicate structures). I found some code for correcting duplicate structures such as https://github.com/jheinly/sfm_duplicate_structure_correction, but they require inputs should include both inlier match data and camera/point cloud information in VisualSfM format. I can get the second one from COLMAP model converter but it seems there is no converter for feature matching.

Do you think there is any workaround or do you think it is good idea to integrate this functionality into COLMAP in the future? 

[Johannes Schönberger]

Hi,

Sorry for the slow answer. Haven't tested this in a long time, but this should be useful for you: https://github.com/colmap/colmap/blob/dev/scripts/python/export_to_visualsfm.py

Cheers,

Johannes