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

generate transform info with low-resolution than apply to hight-resolution #89

Open
qq297110281 opened this issue Jul 7, 2020 · 1 comment

Comments

@qq297110281
Copy link

this project really help me a lot. great work! recently i trying to make it work on arm device(which is Low frequency CPU ),i want to generate the transformations (first pass)with 1280x720 gray YUV images, and then apply (second pass)to the origin YUV images (3840x2160). can it work properly ?

@georgmartius
Copy link
Owner

Hi, this can work, but I did not think about it too deeply. If all related transforms are scaled it might work out of the box.
The transform file contains entries of the type: (LM dx dy fx fy fsize contr match). dx,dy are detected movements of a small field defined by fx,fy and size fsize. If you only scale dx,dy,fx,ft (in your case multiplied by 3) it might work.

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

2 participants