@Gargron Linux has a package called tesseract for OCR. I tried making an auto translator bash script for document scanning with it once https://www.bitchute.com/video/yYyZWUtIEuG5/. It's ok. Don't even bother trying anything without Latin characters.