Take the 2-minute tour ×
Programmers Stack Exchange is a question and answer site for professional programmers interested in conceptual questions about software development. It's 100% free, no registration required.

word extraction illustration

My Situation

I need to develop an application which does crowd-sourced Image-To-Text service (just like google's reCaptcha). So I need an OCR API which can find words in an image with some text.

My Research

  • I have gone through JavaOCR Library.
  • Also Tessaract (which is not implemented in Java & Tess4J Wrapper doesn't have anything I'm looking for).

So basically I have nowhere to go, so I need help.

What I want

  • Is there any API (preferably in Java) that does exactly as I have shown in the image above?
  • If no such API exists, is there any EXACT algorithms suitable for this process?
  • If not, what's your suggestion to achieve this?

Note: I don't want another just image-to-text APIs. I need an API which finds WORD LOCATIONS in the input image.

share|improve this question

closed as not a real question by Giorgio, Glenn Nelson, Dynamic, World Engineer, Martijn Pieters Feb 20 '13 at 11:17

It's difficult to tell what is being asked here. This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form. For help clarifying this question so that it can be reopened, visit the help center.If this question can be reworded to fit the rules in the help center, please edit the question.

    
If the OCR library does not provide boundaries then you could use it as a validation tool. Find "objects" yourself with a recursive fill algorithm. Drop "objects" you consider to be noise. Pass objects to the OCR library to see if it's a letter. You then have a confirmed letter and the location. Not ideal but... –  mike30 Feb 7 '13 at 19:31

1 Answer 1

I used a product called Prime Recognition to create an application with similar requirements a few years ago. It will give you the positions of every letter found on the page.

It’s far from free (it was low 5 figures at the time), but has a good API and delivered good results.

http://www.primerecognition.com/#sthash.yD85Mr5g.dpbs

share|improve this answer

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