October 18, 2019, 06:25:34 PM

Author Topic: convert scans (images) of pages of a book to text ?  (Read 146 times)

Offline RemiD

  • Hero Member
  • *****
  • Posts: 937
convert scans (images) of pages of a book to text ?
« on: September 21, 2019, 12:26:49 PM »
hi,

i have found a scan of the pages of an old book (pdf) on the web but i can't select the text (it is not protected, i have checked), it seems like scan images in a pdf...

any idea how i can convert these images to (selectable) text ?

thanks
DualCore AMD E-450, 1646 MHz - 6 Go DDR3 1333 SDRAM - AMD Radeon HD 6320 Graphics (384 Mo) - Windows 7 Home Premium - DirectX 11.0

Offline Matty

  • Hero Member
  • *****
  • Posts: 751
    • MattiesGames
Re: convert scans (images) of pages of a book to text ?
« Reply #1 on: September 21, 2019, 12:54:31 PM »
You will need software that does optical character recognition.

It does exist, I've used free software at work.

But you will likely have to correct a lot of it.

Offline Amon.

  • Full Member
  • ***
  • Posts: 140
  • What? There's no ceiling outside?
    • Amon.Pro
Windows 10 Pro - 32GB DDR4 RAM - GEFORCE RTX 2070 8GB - AMD RYZEN 7 8 CORE - WATERCOOLING.

Offline RemiD

  • Hero Member
  • *****
  • Posts: 937
Re: convert scans (images) of pages of a book to text ?
« Reply #3 on: September 21, 2019, 09:48:33 PM »
ok, i was aware of "ocr", will take a look. i was wondering if there was another solution... thanks
DualCore AMD E-450, 1646 MHz - 6 Go DDR3 1333 SDRAM - AMD Radeon HD 6320 Graphics (384 Mo) - Windows 7 Home Premium - DirectX 11.0

Offline Derron

  • Hero Member
  • *****
  • Posts: 2486
Re: convert scans (images) of pages of a book to text ?
« Reply #4 on: September 21, 2019, 10:36:25 PM »
You could even translate via your "translate" app if you have a android phone with google-apps. Open the picture there and get the text translated (and selectable). Might be of interest too.


bye
Ron

Offline RemiD

  • Hero Member
  • *****
  • Posts: 937
Re: convert scans (images) of pages of a book to text ?
« Reply #5 on: September 22, 2019, 07:54:10 AM »
@Derron>>thanks, i will take a look
DualCore AMD E-450, 1646 MHz - 6 Go DDR3 1333 SDRAM - AMD Radeon HD 6320 Graphics (384 Mo) - Windows 7 Home Premium - DirectX 11.0

Offline RemiD

  • Hero Member
  • *****
  • Posts: 937
Re: convert scans (images) of pages of a book to text ?
« Reply #6 on: September 22, 2019, 07:19:00 PM »
so... OCR kinda works, but only if the background of the page is white with uniform lighting and no imperfections... but in my case this does not help.  ;D
anyway, it was interesting to try it.
DualCore AMD E-450, 1646 MHz - 6 Go DDR3 1333 SDRAM - AMD Radeon HD 6320 Graphics (384 Mo) - Windows 7 Home Premium - DirectX 11.0

Offline Derron

  • Hero Member
  • *****
  • Posts: 2486
Re: convert scans (images) of pages of a book to text ?
« Reply #7 on: September 22, 2019, 08:15:54 PM »
Did you try google translate? I used it for signs in Czech on a camping place - random dusk time picture. And it translated well. Used it on menu cards ... worked well. Might work better than traditional OCR.


bye
Ron