I'm good at starting projects but never really completing them. So I think I'll start another one. Back in April 2024, a book was published, for the first time, in English ("Passcode to the Third Floor Secretariat"). So I don't need to use OCR and online translation portals for that one. But another book, "Fully Controlled Areas", ISBN 978-89-90959-28-7, has not been translated to English and probably never will be. (And I have got the only copy I will ever have, trust me). Lots of crude, hand-drawn illustrations of people starving and getting executed in North Korea. The text should be quite revealing. I neither read, speak nor understand Korean (or I would not be making this post).
If I get a flatbed scanner, and hold the book down on it, should I use "Tesseract"? If the latest version, version 5, needs the latest libraries (i.e. won't work with Win7), I can always pull out an unused 500 GB SSD and install the latest Ubuntu. (AMD 8150, 64 GIGS of RAM, Radeon 6770 completely passive-cooled video, that should be enough.)
Using the command line for each page does not bother me. It will be just like using the command line to convert "Corel PhotoCD" files to a JPG. After I get the text files (Korean characters can be viewed in Notepad for Win7, not sure about Office 97/03), then I can drop the text into an online translation web portal.
Does anybody have any experience in this area? The documentation seems to suggest this will recognize text well. Does anyone have experience in this area?
If I get a flatbed scanner, and hold the book down on it, should I use "Tesseract"? If the latest version, version 5, needs the latest libraries (i.e. won't work with Win7), I can always pull out an unused 500 GB SSD and install the latest Ubuntu. (AMD 8150, 64 GIGS of RAM, Radeon 6770 completely passive-cooled video, that should be enough.)
Using the command line for each page does not bother me. It will be just like using the command line to convert "Corel PhotoCD" files to a JPG. After I get the text files (Korean characters can be viewed in Notepad for Win7, not sure about Office 97/03), then I can drop the text into an online translation web portal.
Does anybody have any experience in this area? The documentation seems to suggest this will recognize text well. Does anyone have experience in this area?
Comment