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
window_w32.cpp: remove "gutter" #19308
Comments
@vpisarev could you comment the solution? |
here's a little something you can run that demonstrates the issue: |
if desired, I could extend that code to center the displayed image in the window, and also implement WINDOW_KEEPRATIO for this backend (it's currently acting like FREERATIO, stretching). |
have no objections to accept the patch |
The gutter exists even in opencv_attic repository and looks like some hack for old Windows versions support. The core team discussed the issue and decided that we do not need it any more. Feel free to propose patch that removes extra border. |
I'll get it done tonight. would you prefer pull request against master or 3.4 branch? I see the 3.4.14 milestone tag and thought I'd ask. |
Please target PR to 3.4 branch |
I am considering looking into making I was meaning to run the test I see in the highgui module source... I have told cmake to BUILD_TESTS but the generated VS solution's please feel free to stop me if working on window_w32.cpp is a waste of time. I heard about plans for bigger changes in the highgui module. |
I would like to remove the 1-pixel "gutter" that sits between the displayed image and the window's border in win32
imshow
. the functionicvCalcWindowRect
could either get its gutter constant set to 0 or the entire calculation could be simplified (my preference).opencv/modules/highgui/src/window_w32.cpp
Lines 1119 to 1142 in 941a979
reason: when you send the window into fullscreen, you'd expect the image area to be the precise resolution of your screen, right? with that gutter, it's not. it's smaller by two pixels (window decoration disappears, the gutter stays). a precisely screen-sized image shown will be slightly off, which either causes rows and columns of the picture to disappear (GDI: nearest neighbor) or the whole picture gets resampled and turned into mush (OpenGL: bilinear). the gutter also has no function that I can see.
observe the gray gutter:
my question: would you like the pull request to be against 3.4 or master? I ask because last time I had a pull request against the wrong branch, having to rebase overwhelmed me and I'd like to avoid repeated embarrassment.
The text was updated successfully, but these errors were encountered: