search

kwea123 / nerf_pl

NeRF (Neural Radiance Fields) and NeRF in the Wild using pytorch-lightning
https://www.youtube.com/playlist?list=PLDV2CyUo4q-K02pNEyDr7DYpTQuka3mbV
MIT License
2.74k stars 483 forks source link

COLMAP Error : Mismatch #33

Closed theCodes7 closed 4 years ago

theCodes7 commented 4 years ago

Hello, I have created my dataset and trying to train and test it. I get following error when I run 360 inward-facing scene. AssertionError: Mismatch between number of images and number of poses! Please rerun COLMAP! In my dataset, there are 45 images however when I run COLMAP, it detects 14 images only but in the colmap_output.txt it shows details of 45 images.

O/P of running COLMAP cell Post-colmap Cameras 5 Images # 14 Points (1918, 3) Visibility (1918, 14) Depth stats 9.552434898035012 42.84922232367337 17.069477140501178 Done with imgs2poses

What could be the reason ?How should I try to run on my dataset?

kwea123 commented 4 years ago

I think you need to check your images, make sure they are not too dark/bright, contain enough feature points, and the poses between images shouldn't change too much. Are you running COLMAP with GUI? If so you can check the number of successfully reconstructed images in the gui.

If you are able to rewrite the code, you can try to only extract successful images (14 out of 45 in your example) and use them to train, it is a little more complicated than re-taking clean photos in my opinion though.

theCodes7 commented 4 years ago

I tried a clean dataset and it worked. Thanks!