Colmap preprocessing

[dogyoonlee]

Hi

I wonder how to generate the camera pose for blurry inputs (e.g. blurcozy2room scene).

When I run the colmap with 34 images in blurcozy2room directory, error raise.

ERROR: the correct camera poses for current points cannot be accessed

However, when I run the same code for clean image set, it works.

I guess colmap seems to be having trouble getting it to work correctly.

Could you share how to generate the camera poses for blurry inputs?

Thank you !

[limacv]

We export the GT camera poses directly from the Blender file for the synthetic dataset (as is the cozyroom).

[chenkang455]

We export the GT camera poses directly from the Blender file for the synthetic dataset (as is the cozyroom).

how about the real-blur dataset?

[limacv]

We extract from COLMAP (which is why it might be inaccurate and we don't use it for numerical evaluation)