-
Notifications
You must be signed in to change notification settings - Fork 5
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
test suite; generalize past 8x8 #2
Comments
Is this right?
|
Right. |
Good! I'll build that. But instead of Lena, Jelly Beans http://sipi.usc.edu/database/download.php?vol=misc&img=4.1.08 (smooth gradients) and Mandrill http://sipi.usc.edu/database/download.php?vol=misc&img=4.2.03 (high frequency fur). By now, major journals are refusing articles that use Lena, https://en.wikipedia.org/wiki/Lenna#Criticism. |
Committed test cases for 8x8. |
Build: Replace CFLAGS = -O3 -Wall -W to CFLAGS = -O3 -Wall -W -std=c++11 |
What's your question about |
https://ffmpeg.org/ffmpeg-filters.html#deblock allows block sizes from 4 to 512 with a default of 8. Although ffmpeg is related to avisynth, its deblock filter didn't come from Costella, but from "Ramkishor, K., Karandikar, P.: A simple and efficient deblocking algorithm for low bit-rate video coding. In: IEEE Inernational Symposium on Consumer Electronics (December 2000).". That filter would provide another way to test and compare the performance of this repo. |
Github user zvezdochiot suggests block sizes other than 8x8.
This change would be invasive, because 8x8 is hardcoded in loops such as costella/costella_unblock.c lines 1060-1300, 1500-1750, 2020-2170, 2260-2400, 2590-2780... .
This would first need some test cases to ensure that all the 8's were caught.
The text was updated successfully, but these errors were encountered: