Author Topic: Tesseract-OCR  (Read 6738 times)

0 Members and 1 Guest are viewing this topic.

Offline Israel Vega

  • Newbie
  • *
  • Posts: 22
  • User-Rate: +0/-0
Tesseract-OCR
« on: May 21, 2013, 01:47:33 AM »
Somebody have used Tesseract-OCR with PowerBasic?

I am using with my application in PowerBasic but from command line with:

Code: [Select]
ShellExecute(ByVal %Null,"", "c:\tesseract-ocr\tesseract.exe","c:\tesseract-ocr\image.tif c:\tesseract-ocr\text.txt" , ByVal %Null, %SW_HIDE)
And I'm surprised with this great OCR. But I not know how include API's in PB.
If someone already has experience with this and can post a sample.

This is a link of API's but is in language C:

https://code.google.com/p/tesseract-ocr/source/browse/trunk/api/baseapi.h

« Last Edit: May 21, 2013, 01:52:08 AM by Israel Vega »

Online José Roca

  • Administrator
  • Hero Member
  • *****
  • Posts: 2522
  • User-Rate: +209/-0
  • Gender: Male
Re: Tesseract-OCR
« Reply #1 on: May 21, 2013, 02:54:57 AM »
C++ classes can't be used with PowerBASIC.

Offline Israel Vega

  • Newbie
  • *
  • Posts: 22
  • User-Rate: +0/-0
Re: Tesseract-OCR
« Reply #2 on: May 21, 2013, 05:56:03 PM »
Thanks Jose...I'll keep looking for a good OCR...Yesterday I was test ABBYY OCR finereader and is Excelent but I no found a interface
for development.  This have better precision that Tesseract.  Also if somebody know of a good OCR...free or commercial version.



« Last Edit: May 21, 2013, 07:14:51 PM by Israel Vega »

Offline Mike Stefanik

  • Full Member
  • ***
  • Posts: 104
  • User-Rate: +10/-6
  • Gender: Male
Re: Tesseract-OCR
« Reply #3 on: May 23, 2013, 08:13:01 AM »
http://www.nicomsoft.com/products/ocr/

I've never worked with it, but they have both a standard DLL and COM interface and it looks like they support a number of different languages (not PowerBasic though, so you'd have to create the declares for it). Just browsing the documentation, it looks like most of their DLL functions just use handles and integral data types so I imagine converting the C prototypes to PB should be fairly straight-forward.


Offline Theo Gottwald

  • Administrator
  • Hero Member
  • *****
  • Posts: 1058
  • User-Rate: +30/-4
  • Gender: Male
    • it-berater
Re: Tesseract-OCR
« Reply #4 on: May 26, 2013, 06:36:54 AM »
Does anybody know an OCR-Kit that can also "Screen-Read" Smoothed Fonts
and the new "Sub-Pixel" ("Cleartype") Fonts" on Vista, W7 and Win 8 ?
All such stuff i have seen can read only b/w Fonts and fails on Cleartype.

About "Nicomsoft". The price is >$1000 therefore i'd ask them to deliver a PB Header.
Its their advantage to make one.
« Last Edit: May 26, 2013, 06:39:50 AM by Theo Gottwald »