C/C++/Objective-C text recognition library

Question

Does anyone know of any free/open-source text recognition libraries in C/C++/Objective-C? Basically something that can scan an image, and read out all of the plain text.

Solution 1

The most famous one is Tesseract OCR developed initially by Motorola and later become open source. It is also promoted by Google.

There are a few more, perhaps not as famous as Tesseract: http://en.wikipedia.org/wiki/OCRopus http://jocr.sourceforge.net/

Solution 2

Open Source OCR Library: 1.ABBYY Mobile OCR Engine is a powerful software development kit which allows developers of mobile and small footprint applications to integrate highly accurate optical character recognition (OCR) technologies that convert images and photographs into manageable and searchable text. Toolkit supports the most popular mobile platforms and devices - iOS (iPhone) and Android. http://www.abbyy.com/mobileocr/

2.Tesseract is probably the most accurate open source OCR engine available. https://code.google.com/p/tesseract-ocr/

Solution 3

Here is an interesting article on how to extract numbers from picture.

The author collects large number of samples to train his neural networks algorithm and I believe doing the image processing with OpenCV.

Comments

[Nubzor]

After the jump it mentions HP as a source ...

[Jonathan]

Tesseract is very good, however beware of it's limitations when it comes to dealing with non-latin character scripts (ie chinese).

[Mathai]

Why did this get downvoted?

[RyanfaeScotland]

I'm not a down voter but it did read a lot like an ad for ABBY Mobile without an initial disclaimer that Sumit was associated with them. I don't think this is the situation but I could see a trigger happy reader down voting as a result.

[James Wierzba]

I tried tesseract out with a few sample and it recognizes a few letters out of entire paragraphs and the rest is junk symbols, even when I type in english whole words in notepad and screenshot it