Ruby + OCR

Discussion in 'Ruby' started by Zbigniew Kowalski, Nov 17, 2011.

  1. hi!
    I need to create sort of service that will OCR a simple image of a
    document send via e-mail to serwer.

    Would you recommend any worth-to-use libraries to do so?

    Thank you.
    Z
     
    Zbigniew Kowalski, Nov 17, 2011
    #1
    1. Advertisements

  2. Zbigniew Kowalski

    Fester Guest

    Hello,

    Recently I was digging into the same area in Python and came to
    following conclusions:
    1. You must choose between pretty expensive and proprietary Abbyy
    command line OCR SDK and free Tesseract OCR. Abbyy's product is great
    in recognition, but have very limiting license, while Tesseract is
    great and trainable, but have very poor layout analysis.
    2. I am not aware about any existing wrapper over either of these
    products. Writing a basic wrapper won't be a real problem though,
    since basic interaction with them is limited to forking an external
    process. Additionally, Tesseract has an API bindings for Python, it
    seems that implementing them for Ruby would be an easy task too.

    Tesseract would work for you if you have an evenly formatted amounts
    of text. Otherwise you would have to implement image layout analysis
    engine on your own. Also, you would better use SVN trunk of Tesseract,
    because it contains many changes comparing to the last packaged
    version.
     
    Fester, Nov 21, 2011
    #2
    1. Advertisements

  3. Hi,
    Thank you for this info! I'm very appreciated. It will be very useful.
    I guess it won't be a great problem because as far as I saw the sample
    pages - text use a clear type fonts (most fixed-width) - like old
    fashioned typewriters. :)

    Regards
     
    Zbigniew Kowalski, Nov 21, 2011
    #3
  4. Zbigniew Kowalski

    Fester Guest

    Hello,

    I think it will be fine unless you have big amonunts of noise. Also
    you can check out Tesseract's training abilities, and probably even
    enforce a better recognition of a well-known font face by retraining
    the OCR.
     
    Fester, Nov 22, 2011
    #4
  5. Zbigniew Kowalski

    nikolaykhl Guest

    Hello there,
    My name is Nikolay Khlebinsky, i work @ ABBYY.

    As Fester mentioned, ABBYY provides the most accurate OCR (for
    example, have look at http://www.splitbrain.org/blog/2010-06/15-linux_ocr_software_comparison).

    We are currently launching a new cloud-based OCR SDK suitable for
    small businesses and single developers. It has a well-composed
    developer guide, a good set of sample codes (including python) and
    it’s free for the testing period

    Would you like to participate in the closed beta testing program of
    our OCR SDK?

    All you have to do to participate is sign up at www.ocrsdk.com, fill
    in a short form a start developing your application. Please fill the
    “Where did you hear about ABBYY Cloud OCR SDK?” field with “Nikolay
    invite”.

    Feel free to contact me if you have any questions.
    Best regards, Nikolay Khlebinsky.
     
    nikolaykhl, Nov 23, 2011
    #5
  6. Zbigniew Kowalski

    Nikolay Guest

    Hello there,

    My name is Nikolay Khlebinsky, i work @ ABBYY.

    As Fester mentioned, ABBYY provides the most accurate OCR (for
    example, have look at http://www.splitbrain.org/blog/2010-06/15-linux_ocr_software_comparison).

    We are currently launching a new cloud-based OCR SDK suitable for
    small businesses and single developers. It has a well-composed
    developer guide, a good set of sample codes (including python) and
    it’s free for the testing period

    Would you like to participate in the closed beta testing program of
    our OCR SDK?

    All you have to do to participate is sign up at www.ocrsdk.com, fill
    in a short form a start developing your application. Please fill the
    “Where did you hear about ABBYY Cloud OCR SDK?” field with “Nikolay
    invite”.

    Feel free to contact me if you have any questions.
    Best regards, Nikolay Khlebinsky.
     
    Nikolay, Nov 23, 2011
    #6
  7. Zbigniew Kowalski

    Nikolay Guest

    Hello there,

    My name is Nikolay Khlebinsky, i work @ ABBYY.

    As Fester mentioned, ABBYY provides the most accurate OCR (for
    example, have look at http://www.splitbrain.org/blog/2010-06/15-linux_ocr_software_comparison).

    We are currently launching a new cloud-based OCR SDK suitable for
    small businesses and single developers. It has a well-composed
    developer guide, a good set of sample codes (including python) and
    it’s free for the testing period

    Would you like to participate in the closed beta testing program of
    our OCR SDK?

    All you have to do to participate is sign up at http://www.ocrsdk.com,
    fill in a short form a start developing your application. Please fill
    the “Where did you hear about ABBYY Cloud OCR SDK?” field with
    “Nikolay invite”.

    Feel free to contact me if you have any questions.
    Best regards, Nikolay Khlebinsky.
     
    Nikolay, Nov 23, 2011
    #7
  8. @Fester: Most of the images would be taken with cheap cameras (in the
    beginning ) - so I guess it need to be trained to eliminate camera
    noise and "shakes".

    @Nikolay:
    I would like to participate to the beta test to compare which solution
    would be more suitable for my idea. I will send the info this weekend.

    Thanks and regards
    Z
     
    Zbigniew Kowalski, Nov 25, 2011
    #8
  9. Zbigniew Kowalski

    nikolaykhl Guest

    Hi Zbigniew,

    I haven't seen you signing in for the beta testing program. Did you
    face any difficulties with registration?
    Feel free to contact me if you have any questions.

    Best regards, Nikolay.
     
    nikolaykhl, Nov 28, 2011
    #9
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.