If I’ll process this picture. How OpenCV will detect there are two different pages and the output will be Two files of each page.
Building such an application requires you to put together a pipeline of image processing steps.
Is there something I’m particular that you’re struggling with?
I would start by trying to segment the paper from the background by searching for a bright box.
A good place to start is ‘edge detection’ and ‘Hough lines’ and ‘binary thresholding’.
Once you have that figured out you can start looking into separating the 2 pages.