Tell me more ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.
up vote 23 down vote favorite
9

Something I've had an occasional need for, but I've never found an application I liked - OCR

Basically I want to take a photo/scan of a document, and convert it to a text document of some kind (Ideally an option for plain-text, perhaps a .doc or .pages)

Requirements:

  • Must have a native (Cocoa) GUI, not under X11
  • Free

Optional pluses:

  • Doesn't require installation, just drag-app-to-Applications-folder (a lot of the OCR utilities I found required libraries to be installed and such)
  • Support images in scanned documents
  • (Apple-)scriptable
  • Open source
share|improve this question
1  
I'd love a program that adds text to scanned PDF documents. (To get some "text layer" behind the scanned image, to be used to search the document and to select text.) I know some scan software (like for Canon multiple purpose printer/scanners) does it on the fly, but most PDFs are simply images without virtual text. Maybe some day a solution will posted at "Software to help me go paperless on a Mac" at superuser.com/questions/7969/… – Arjan Jul 21 '09 at 9:14
2  
Haha, nice workaround: use Google to convert your scanned PDFs to text: labnol.org/software/… – Arjan Jul 21 '09 at 9:16
Sorry to keep commenting, but as it is not a direct answer to your question: more online options such as Scribd iPaper and Evernote in "The Best Software Tools for the Paperless Office" at labnol.org/software/organize/… – Arjan Jul 21 '09 at 9:33
@Arjan: I have a CanoScan 5000F and it adds the text to the scanned PDF documents. If I remember correctly it is actually Omnipage SE (working in the background) the provides this functionality. Anyway, I love this feature. – Ludwig Weinzierl Jul 21 '09 at 9:38
Thanks for reminding me about that name. Indeed, the Canons I used shipped with Omnipage SE. But I never figured out how to use Omnipage to achieve the same without human intervention (like using Omnipage to create searchable PDFs with other scanners). I know that without human validation the text in the PDFs created by the Canons is not 100% accurate, but it's close enough. (Sometimes, when one forgets about the errors that OCR can make, one might wonder why some text or sentence cannot be found while it's clearly in the scanned image. Or why the copied text differs from the shown image.) – Arjan Jul 21 '09 at 9:50

closed as off topic by KronoS, Journeyman Geek, random♦ Apr 2 at 0:07

Questions on Super User are expected to relate to computer software or computer hardware within the scope defined in the FAQ. Consider editing the question or leaving comments for improvement if you believe the question can be reworded to fit within the scope. Read more about closed questions here.

9 Answers

up vote 7 down vote accepted

When I researched this topic a while ago there wasn't any free software for any platform that produced reasonable quality output.

The Optical character recognition article at Wikipedia lists the following free OCR applications:

I only tried gocr from these, it has no gui and the qualitiy of its output is very low.

I's suggest to go with a commercial product. Either ABBY Finereader or OmniPage, both of which have OS X versions. They are often bundled with scanners and you can buy them pretty cheap if you don't need the latest version.

share|improve this answer
1  
gocr certainly fails badly on scanned lab-book pages. Even after diddling the density and contrast, bits of the underling graph show through and confuse the program... – dmckee Jul 21 '09 at 14:03
up vote 5 down vote

Tesseract, is free, has an OSX port and a how to

share|improve this answer
I've just used it, it works. Installed it with ports. Be aware that you have to have tiffs with the extension being ".tif" with 8 bit color depth or less and no alpha channel. – J. Pablo Fernández Jul 18 '10 at 19:49
up vote 4 down vote

Never heard of any good free OCR for Mac :] There is GOCR, but it is rather crappy. From the low cost apps I would recommend VelOCRaptor. You can try it out for free.

share|improve this answer
Sadly, VelOCRaptor only works up to OS X 10.6. – mklement Mar 19 at 13:51
up vote 2 down vote

how about evernote? send them the image, they ocr it for you.

share|improve this answer
1  
Note that EverNote OCR only works for finding text -- it does not give you access to the converted text. – Rubistro Oct 11 '12 at 2:25
up vote 2 down vote

Sorry, but some command line action is involved in this solution...

If all you need to do is convert a PDF containing scanned pages into text, the following method has given me good results - where GUI based tools such as VelOCRaptor have failed (I'm talking about a 134 page PDF doc with scanned pages).

All programs in the tool chain are free or come with OSX.

  • With Preview, save the PDF as TIFF, 150dpi. Make sure you have enough disk space to play with.
  • Run the TIFF through Tesseract (install using MacPorts / Fink)
  • Now you have a raw text file, which can be spell checked using any good editor (TextWrangler, TextEdit, etc.)

Good luck!

share|improve this answer
up vote 1 down vote

I have just found something called 'PDF OCR X Community Edition. It's pretty basic - it just gets plain text out without formatting. However, it works quite well. I used it for scanning German, even though officiall it only works with English, and it was ok.

share|improve this answer
up vote 1 down vote

I haven't been able to find anything for free, but PDFScanner is cheap at $15, is a native OS-X app (Snow Leopard and Lion only, AFAICT), and both scans to an OCR'd PDF and lets you open and OCR an existing PDF. It's the only not-horribly-expensive thing I've found.

share|improve this answer
1  
He did specify 'Free' as a requirement, didn't he. How can this answer be valid then? – František Žiačik Oct 23 '12 at 21:32
up vote 0 down vote

http://www.thefreecountry.com/utilities/ocr.shtml

Sounds like if you have microsoft office, there is a tool that can convert images into files. Other than that, i am not seeing anything that is free for os x.

http://discussions.apple.com/thread.jspa?messageID=9807438 Very recent apple discussion in the support forums

you did mention you only want a native cocoa app, if you could consider some of the linux builds, you might have some luck as there are a few options there.

share|improve this answer
up vote -1 down vote

I know it's not free, but I've had reasonably good experience using ReadIris, which you can find for around $60. Basically it will take scans, jpegs or several other image formats, or even PDFs containing scanned data, do OCR on them, and write as PDF ("searchable" -- i.e., text and/or the image itself).

share|improve this answer
1  
He did specify 'Free' as a requirement, didn't he. How can this answer be valid then? – František Žiačik Oct 23 '12 at 21:31

Not the answer you're looking for? Browse other questions tagged or ask your own question.