I have an image to process: a scan of a book page. I want to preserve the useful info including the page numbers in the corners, but to crop the junk and scan artefacts on the edges of the image. What’s the best solution (preferably in Python) to achieve that?
I need either a complete solution to trim away the garbage or at least a way to detect where text block edges start— some kind of OCR library to detect dimensions of the useful information area in a page scan.
Thanks
UPD: Sample: https://i.stack.imgur.com/k5kvk.png I want to be able to crop the good part of the scan, so that there is not part of another page from across the bind, no black outside the pages, no distortions on the edges.
