Is there any way to make use of JPEG's shrink-on-load feature to speedup the loading of huge images in OpenCV, when we don't need the full resolution?
imread() does not seem to provide options apart from color management.
I'm surprised I haven't been able to find any discussion or feature request around this.