search

biomedia-mira / drop2

drop2 - intensity-based image registration
Other
36 stars 3 forks source link

Parameters for registering a perspective transformed image #10

Closed WhaSukGO closed 2 years ago

WhaSukGO commented 2 years ago

First of all, thank you for such a marvelous work!

I have a question regarding the parameters.

I'm trying to register 2D perspective-tranformed images to a given template.

For example, in the image below, I'm trying to register image on the left to template (which is on the right)

Example (The image is from here)

No extra transformation is needed, but only homography is required.

Is it possible to estimate the parameters of homography via this project? If so, could you provide some tips on how to set the parameters?

Thank you!

bglocker commented 2 years ago

Thanks for the kind feedback.

I'm afraid drop2 has no projective transformation model implemented. The linear registration models are currently limited to rigid, similarity, and affine. As drop2 is primarily used in the context of medical image registration, there has been no need for projective transformations, so far.

WhaSukGO commented 2 years ago

I see. Thank you again! By the way, may I ask if there's any algorithm for registering projective transformed images?

bglocker commented 2 years ago

I would suggest to check out OpenCV