Need suggestions for OCR software

August 22, 2008 8:50 AM

Any good free (or really inexpensive) OCR software out there? I looked through the archives here but couldn't find any info that was quite right for my needs.

I have a typewritten manuscript of 300+ pages that I want to suck into Word via OCR. I'm running Windows XP on a Thinkpad. I have a Canon scanner that supposedly has OCR software but it simply sucks, and I want something better. But, you know, free -- or almost free. Googling OCR hasn't helped me much and I'd like some real-world advice from the Green.

Thanks in advance!

And really, I had nothing to do with the whole Bigfoot thing. Swearta god.

posted by Guy_Inamonkeysuit to writing & language (20 comments total)

FineReader by ABBYY is far and above the best out there, all the free ones are really bad.
posted by zeoslap at 9:01 AM on August 22, 2008

The only one that did a half decent job was Cuneiform
posted by zeoslap at 9:05 AM on August 22, 2008

The only free one I mean.
posted by zeoslap at 9:06 AM on August 22, 2008

Also XP has a built in OCR thingy MODI
posted by zeoslap at 9:07 AM on August 22, 2008

You could always just scan it to PDF or TIFF yourself and send those files to someone who's already paid for non-sucky OCR software. 300+ pages could be a lot of work, but some people are inexplicably helpful.

I have had great success with ReadIris PRO 11, but it's $129.
posted by onshi at 9:26 AM on August 22, 2008

As a datapoint, I have sucessfully used OneNote to OCR some scanned pages
posted by Memo at 9:29 AM on August 22, 2008

I have used ReadIris before, onshi -- but I had a free version I'd gotten form somewhere and that was on another computer. I liked it. I've been looking at it but my budget doesn't like the cost.

I looked into MODI a few weeks ago, zeoslap, but it isn't on my machine, which I bought second-hand and for which I do not have the XP install disks.
posted by Guy_Inamonkeysuit at 9:51 AM on August 22, 2008

Sorry if this is a derail, but are there any online services that let you upload a PDF/TIFF and then download a text file, preferably for free (maybe with a registration)? Sounds like something Google should be doing, since I imagine they're already working on OCR as part of their world domination scheme.

Wasn't there a project where you'd take a picture of a street sign or the name of a restaurant/product, and then text that to Google and then they'd give you relevant info on it? Or was I just dreaming? If not Google, ChaCha could probably handle that if they can afford the bandwidth.
posted by mccarty.tim at 11:07 AM on August 22, 2008

Tesseract
posted by i_am_a_Jedi at 11:37 AM on August 22, 2008

FineReader is a must.
posted by zouhair at 1:07 PM on August 22, 2008

FineReader at #399.00 is much too expensive. ABBY's Scan to Office might do the trick, though; lots cheaper.
posted by Guy_Inamonkeysuit at 1:30 PM on August 22, 2008

Just to clarify, MODI does not come with Windows XP. It's a part of Office. Do you have your Office install discs?
posted by phrayzee at 1:47 PM on August 22, 2008

Seconding Tesseract. It's all command-line only and takes some experimenting to set up, but once it works it does the job very well.
posted by husky at 1:53 PM on August 22, 2008

A typing service (for example) wouldn't cost you any more than finereader costs. So maybe you should consider doing that.

What you should under no circumstances consider doing is downloading FineReader for free from any popular BitTorrent website even though it is available there. That would be stealing.
posted by Authorized User at 2:53 PM on August 22, 2008

SimpleOCR
posted by nedpwolf at 3:32 PM on August 22, 2008

2nding SimpleOCR. Used it to import some really old documents, and it worked wonders.
posted by deezil at 4:06 PM on August 22, 2008

I think I will give SimpleOCR a shot, based on what I read there and the recs from nedpwolf and deezil. Thank you all, as always your knowledge rocks the Casbah.
posted by Guy_Inamonkeysuit at 4:37 PM on August 22, 2008

Um, adobe reader can do OCR. I know this may be hard to believe (if I had been a gum chewer, I'd have swallowed my gum), but you open a pdf in it, and you can highlight the text and copy them into another program like textedit or word.
posted by history is a weapon at 11:30 AM on August 28, 2008

Um, adobe reader can do OCR. I know this may be hard to believe (if I had been a gum chewer, I'd have swallowed my gum), but you open a pdf in it, and you can highlight the text and copy them into another program like textedit or word.

Only applicable if said pdf contains text. If it is composed solely of images, you won't get very far this way.
posted by i_am_a_Jedi at 4:11 PM on August 29, 2008

Okay, after extensive asking around and what-not, I have downloaded SimpleOCR and it seems to be sufficient for my needs. nedpwolf and deezil, you guys win the enchilada. Thank you!
posted by Guy_Inamonkeysuit at 10:37 AM on September 2, 2008