Unsolved QImage column first support
-
Hi,
I am wondering if Qt team has a plan to support column first QImage. To convert from column first data (naturally acquired by column based scanning) to traditional row first QImage is not a quick operation but assume when QImage data is used elsewhere (such as rendering), it will be as easy as row first or column first.
Or if there are other solutions, please suggest.
Thanks,
Patrick. -
@zxpatric said in QImage column first support:
Or if there are other solutions, please suggest.
Return an image for Qt::DisplayRole
-
Hi,
This forum is not the right place to talk about what the Qt folks are going to implement as it is a user forum.
That said, do you have an example of such an image ?
Most of the acquisition system I saw usually provide low-level data like a Beyer matrix that you have to convert to something usable in terms of image processing if they do not already provide raw RGB or YUV data. -
@Christian-Ehrlicher Can you elaborate a bit further? if I have a buffer with image raw data in column first memory alignment, how do I construct a QImage without expensive transpose operation?
Qt::DisplayRole 0 The key data to be rendered in the form of text.
-
@SGaist Imagine there is a vertical sensor and it scans while rotating around an parallel axis. The data coming out of the sensor are genuinely in the form of column first. Image transposing is not a cheap operation and I assume on the rendering side, it is just a matter of switch the two for loops if rendering is a naive iteration or column-first vectorization instead of row-first vectorization.
-
I worked with a vertical linear sensor, however, the camera using that sensor was generating the image as a Bayer matrix rather than sending data column by column.
-
@SGaist Do you know which forum is for feature request towards Qt developers?
-
None, forums are not for feature request.
You can do that on the bug tracking system.
Some recommendations:
- be precise
- give as much details as possible
- provide references to support your use case
-
@SGaist Thanks a lot!