It’s easy if the split is always at the same point and you can define it using coordinates. And if the document was scanned flat. But in the real world each scan is different and a book scanner photographs curved pages so as not to damage the book and to be fast (i.e. no need to flip the book to flip pages). And recognizing the fold via image processing, splitting, reducing the colour shifts brought on by the fold and then deskewing the split images to compensate for the curve of the page is far from trivial. It’s not impossible, but beyond my capabilities using libraries such as the one you posted, and as far as I know, nobody has yet released an image processing library to do it automatically
2 Likes
This topic was automatically closed after 5 days. New replies are no longer allowed.
