Skip to content

Conversation

@Jordan-Pierce
Copy link

Per the paper, DA3 can take as input camera intrinsics and extrinsics as opposed to estimating them directly. The PR provides two helper functions in read_write_model.py:

  • extract_intrinsics_extrinsics_from_colmap: extracts the intrinsics and extrinsics from the COLMAP model, which is provided from the existing read_model method
  • get_image_paths: gets the absolute image file paths given the COLMAP model (which stores the actual images two directories above where cameras, images, and points3D are located (see here)

It also provides a notebook colmap_extraction_demo.ipynb that demos the methods and the resulting prediction intrinsics and extrinsics with and without the input COLMAP intrinsics and extrinsics.

With COLMAP:
with

Without COLMAP:
without

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant