opencv / opencv/opencv

estimateAffinePartial2D without scaling

Open
#26,329 2 comments 5 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

feature
Dominant language
C++
Stars
90.9k
Forks
57k
Avg merge
2d 15h
Merged PRs (30d)
112

Description

Describe the feature and motivation

I use estimateAffinePartial2D function to stabilize a set of images of a fixed camera dedicated to coastline monitoring. The camera zoom is fixed. There are small camera movements due to for example the thermal dilation of the structure on which the camera is installed.

The stabilization result with estimateAffinePartial2D is quite good, but it could be really better if scaling factor was not computed and left to 1, as it would better suit the camera configuration.

It would be great to have a boolean parameter in estimateAffinePartial2D, in which the user can specifiy if scaling factor has to be computed or not. Otherwise, it might be more simple to implement an other function called for example 'estimateRigidBody2D' that would compute a rigid body (translation + rotation) transformation from matching points.

Additional context

No response

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

Start with the estimateAffinePartial2D API and its documentation link. Determine how to expose an option that fixes scaling at 1, or whether a separate rigid-body transform entry point is needed. Done means matching points can produce translation and rotation without estimating scale, with the API behavior documented and tested.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
computer-vision
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.