/lit/ - Literature

so after three days of continuous tinkering, i finally have a decent workflow for converting book scans into PDFs that are readable on e-ink screens. Pic rell is a picture of an original scan, and the cleaned up version for e-reading
if you take a normal scan of a book and try to read it on an e-reader, it shows up really weird and the background is grey and blotchy. I found a way to isolate just the text and add a plain white background so now, its a plain, sharp, clear black and white page. Its the closest thing to actually owning the book itself in terms of digital reading.
originally I was going to convert them to djvu but all the utilities and tools for djvu manipulation are so outdated that they wont run on a modern system (well, they probably could, but that requires knowledge of the dark arts of UNIX shell scripting that are beyond my grasp). PDF is a pretty universal format and I can compress a whole lot, since its just black and white text + color cover image. I can also set the DPI to be just as sharp as printed text (assuming your e-reader's PPI is high enough to handle it), Also I now easily add (surprisingly accurate) OCR too, so the PDF will be text-searchable.
Im mainly posting this because I'm wondering if there are any other anons interested, I was thinking about doing a write-up of the process