Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Question: Use more than 4 points for pose computation #333

Open
s-trinh opened this issue Jun 5, 2024 · 3 comments
Open

Question: Use more than 4 points for pose computation #333

s-trinh opened this issue Jun 5, 2024 · 3 comments

Comments

@s-trinh
Copy link
Contributor

s-trinh commented Jun 5, 2024

Sorry, this is not a bug but rather a question.


I am wondering if using not only the 4 corners of the marker but some specific (depending on the marker ID) additional corners from the pattern itself would bring more accurate pose estimation?
For instance, when processing time is not a constraint, a first estimation of the marker pose could be computed as usual and then refined (orthogonal_iteration()?) using some additional points.

image

I would say maybe in theory this should bring more accurate pose estimation, but in practice this should be useless/negligeable wrt sensor noise, image distortion, size of the marker in the image, etc.
I think that the 4 corners coordinates are already computed from contour points line fitting, so using additional corners from the pattern should not bring additional useful information?

I am interested about other opinions.

@mkrogius
Copy link
Contributor

mkrogius commented Jun 5, 2024

I tried something like that but it didn't seem to give improvement, at least for far away tags. My guess is that the edge fitting already makes use of most of the useful information for pose.

Some previous discussion here: #19

@s-trinh
Copy link
Contributor Author

s-trinh commented Jun 6, 2024

Thanks for your feedback.
Maybe you could add a specific label for questions or enable Github discussion tab?

@kke19
Copy link

kke19 commented Jun 25, 2024

okay, thank you

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

No branches or pull requests

3 participants