Closed denred0 closed 1 year ago
Yes in this case you can use our integration with hloc to run triangulation: python -m pixsfm.refine_hloc triangulation -h
. You have to provide a reference SfM model in colmap format. You can easily create this reconstruction (holding images, poses and camera intrinsics) using pycolmap. Use hloc to extract and match features, then run triangulation using pixsfm.
Thanks for your work! I am a new in this sphere and try to understand a lot of things. May I use the pixel-perfect-sfm for creating a point cloud with absolute distance to a camera? I have images, camera poses for every image and camera intrinsic.
Thank you!