1

I recently bought an ebook online (DRM-free). It was supposed to be in epub format, which it technically is, but as it turns out they just packaged an image of each individual page rather than the actual text. As result the whole thing is around 100MB and does not work on any of my readers. The only way to read it at the moment is to extract it and use an image viewer.

How can I convert this to actual text to make it usable?

Anthon
  • 7,641
  • 3
  • 23
  • 56
erlking
  • 111
  • 2

1 Answers1

1

What you need to do is run OCR (Optical Character Recognition) on the images of the book.

If you cannot find a program that does the OCR directly on the ePub file, you can try to convert it to PDF first, or extract the images. Make sure that if you do a conversion, that the images do not loose quality, as that would negatively influence the results of the conversion. Programs like Calibre allow you to do such conversions.

There are some online OCR services, mainly for conversion of PDF, make sure that you understand the implications of using such a service, as there is no guarantee that your book information is removed after conversion (and if you images are watermarked, or contain metadata pointing to you, that could get you in trouble).

Anthon
  • 7,641
  • 3
  • 23
  • 56
  • Do you have a recommendation for OCR software? – erlking Oct 10 '16 at 11:02
  • I use tesseract on Linux – Anthon Oct 10 '16 at 13:17
  • To extract the images, you can use a program like PeaZip to uncompress the EPUB (it's basically just a zip file) and then hunt around the extracted folders for the image files. For OCR, you can also use Tesseract on Windows and some other platforms (https://github.com/tesseract-ocr/tesseract/wiki). – thinkulum Oct 12 '16 at 13:18