create lossless PDF files from page images and HOCR files #2
Description
ideally i want to create "lossless PDF files", without the full page images you normally have in scanned PDFs
when you buy books in epub format from "normal" book publishers, then epub is a lossy format: you lose the print formatting, and the images usually have a bad quality (72dpi?).
this is why i buy books in print format, remove the binding with a guillotine cutter, and send them through my ADF scanner at 600dpi, see also collaborative proofreading of scanned books.
downsides of my solution: printed books are more expensive than ebooks, OCR proofreading is required... ideally i have both book versions: printed book and ebook, then i can use the formatting from the printed book, and the text from the ebook.
upsides of my solution: you can reproduce books near-lossless by printing. (good books deserve to be printed!) (the cheapest binding method is stapling booklets with a block stapler.)
ideally i want to create "lossless PDF files", without the full page images you normally have in scanned PDFs, so you get a (lets say) 5 MB PDF from about 1 GB of scanned images, but that is... work in progress, only few developers work on this.
also posted on reddit:
Is it possible to convert an epub to a PDF, and have the page numbers accurately match up with a physical book?