Anybody know of a decent OCR program for OSX

Discussion in 'Mac Apps and Mac App Store' started by Stelliform, Jan 2, 2004.

  1. Stelliform macrumors 68000

    Stelliform

    Joined:
    Oct 21, 2002
    #1
    Here is the situation. I have a client that wants a scanner for OS X. I figure that isn't a problem, however they want it solely to do OCR. What complicates the problem is that they are used to $100 paperports. In fact they still have a Mac running OS 8 with a paperport attached to it solely for scanning and OCRing. (So they are balking at any solution over $100.)

    I have found a few OCR programs on Apple's page, but the one that I a familiar with (Omnipage) is $500 for a Mac. I am about to tell them it can't be done, but I figured that one of you guys probably has a solution. Thanks!
     
  2. crazzyeddie macrumors 68030

    crazzyeddie

    Joined:
    Dec 7, 2002
    Location:
    Florida, USA
  3. mhar4 macrumors member

    Joined:
    May 7, 2003
    Location:
    London
    #3
    I've used Omnipage Pro X, Readiris 7 and Fine Reader (OS 9). I mainly ocr academic journal articles. None of them do anything really amazing. All usually require page by page reformating, although the results are very useful in the end. It's formatting that catches these programs out, not text accuracy.

    Anyhow I bought Omnipage Pro X in the end.
     
  4. Stelliform thread starter macrumors 68000

    Stelliform

    Joined:
    Oct 21, 2002
    #4
    That is what I was figuring I was going to recommend. They are familiar with paperport and it is extremely easy to use and fairly accurate. (Not to mention cheap.)

    They just assumed that all scanners OCR straight to Word right out of the box. :rolleyes:
     
  5. bousozoku Moderator emeritus

    Joined:
    Jun 25, 2002
    Location:
    Gone but not forgotten.
    #5
    PaperPort, Xerox TextBridge, and OmniPage Pro are all owned by the same company. The latter two are essentially the same application but OmniPage Pro has some extra features like (pathetic) .pdf file output.

    It's unfortunate that one company owns the majority of the market. After using OmniPage Pro for Windows for 6 months, I can't imagine that it's worth $500. The accuracy that ScanSoft advertises wasn't possible with what I was doing, but it reported 98-99.886 % accuracy each time.

    Good luck. Give them a nice chair to cushion the fall when they faint. :D
     
  6. mhar4 macrumors member

    Joined:
    May 7, 2003
    Location:
    London
    #6
    Fortunately, I managed to find a second-hand copy of Omnipage Pro X, so didn't get slugged the British equivalent of $500. I agree that it is not worth that at all, given the amount of time you need to spent working over texts doing reformatting. The pdf output is indeed a waste of time. I export to Appleworks and edit and export to pdf from there.
     
  7. yamabushi macrumors 65816

    yamabushi

    Joined:
    Oct 6, 2003
    #7
    OCR might be a useful feature to add to a future Apple document creation product or to OSX itself.
     
  8. cubist macrumors 68020

    Joined:
    Jul 4, 2002
    Location:
    Muncie, Indiana
    #8
    Who sells that, Caere? They have indeed been buying up all the competition. But historically, OCR has been expensive. There is huge R&D involved.

    On the PC side I used to work for a company where we integrated a package called Recore. I was very happy with Recore: it had a clean API and it was reasonably fast and accurate. I did the formatting myself (I don't think formatting is properly part of an OCR package).

    Why don't you see if Recore offers a Mac OS X library, then write a wrapper in Cocoa? Your client will be willing to pay more for a customized program and everyone will be happy. (It works wonders to display their logo on the screen.)
     
  9. bousozoku Moderator emeritus

    Joined:
    Jun 25, 2002
    Location:
    Gone but not forgotten.
    #9
    Caere is no more. I think their Mac competition was Calera--it was something very similar. Everything is part of ScanSoft now.

    Even if there would be a Recore library for Mac OS X, it's not likely to be trivial regardless of what other pieces might be available as open source. It sounds a bit more than $500.
     

Share This Page