SfM from known poses

Open
#2,345 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
35/100
Issue type
Feature
Clarity
Needs clarification
Activity status
Stale
Tech stack
cpp

Research direction

Start by reading the global SfM reconstruction and pose-prior documentation, especially the paths for GPS priors and known camera poses, then trace where image pose information is supplied. Determine whether known poses can be accepted for global SfM, and document the supported input format, implementation entry points, and tests needed to establish that reconstruction uses the supplied positions and orientations.

Written by the indexing model from the issue text.

Description

Hi, I have been trying to add pose priors to my reconstruction but I would like to know the "best practices". I have both position and orientation information for each image.

The documentation supports the use of GPS prior if embedded in the image metadata. So I tried adding the lat-lon-altitude to the metadata, which somewhat improved the results but I wonder if there is a better way than adding the pose information to the metadata of each image.

If I perform a reconstruction from known poses I can provide the position and orientations from a text file, could I do the same for the global SfM reconstruction? Does it even make sense? I imagine I can't be the first as I imagine most might have wanted to do something similar with SLAM priors.

In the end I registered roughly 2k out of the 3k images I originally had, so the results I now have are encouraging but not sufficient using disk+lightglue matches computed exhaustively.

Dominant language
C++
Stars
6.6k
Forks
1.7k
Avg merge
11h 18m
Merged PRs (30d)
2

Getting set up

This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from openMVG/openMVG

All issues in openMVG/openMVG

Similar issues

More C++ issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.