mapillary / mapillary/OpenSfM

Using OpenSFM with a telescopic lens

Open
#1,045 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
3.8k
Forks
899
PR merge metrics
No merged PRs in 30d

Description

I have a set of images which were taken with a telescopic lens, with a very large focal length. The object I am imaging is several kilometers away.

The images do not have any exif metadata attached to them. When I run with no camera model overrides, the reconstruction is performed fairly well - the relative shape and camera position appear to agree well with predicted results. However, this does not produce accurate measurements in real-world units since the focal length is estimated much too low at ~0.92. Unfortunately, when I run with camera overrides (with the actual focal length), the reconstruction cannot be performed since there are no image pairs recognized.

I fear that some prior being used is starting the reconstructed camera position at an estimated distance of a few meters from the object, when the true camera position should be several kilometers from the object. Is there any prior I can override to provide an initial estimated camera position which is more accurate for my long focal length?

Contributor guide

Open the contributing guide

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.

Research direction

The issue names no source file, test, or entry point. Start by locating the camera override handling and reconstruction initialization, then reproduce the runs with and without overrides using the telescopic-lens images. Done would require a confirmed way to handle the long focal length and recover accurate real-world measurements.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
computer-vision
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.