 |
|
| free OCR programs tested here: |
|
|
By nutela - Posted on May 27, 2007 - 14:51:27 (#21678)
Current version when comment was posted: 0.0.2 |
|
|
http://groundstate.ca/ocr
|
|
| new project by 3dEyes |
|
|
By s_d - Posted on May 5, 2007 - 09:41:29 (#21637)
Current version when comment was posted: 0.0.2 |
|
|
Ge is working on image to text translator based on Neural Networks atm:
http://img214.imageshack.us/img214/7529/myocrtest0df0.png
It has great learning ability
|
|
| kdebug photo's->GOCR-> crap :-( |
|
|
By nutela - Posted on May 5, 2007 - 07:46:44 (#21636)
Current version when comment was posted: 0.0.2 |
|
|
I tried to let gocr save kdebug pictures to text but the result is really bad, is it using an internal dictionary? Is there a way to let it learn stuff or display characters it was not sure with in red colour?
Colours could be bound to probability values :-)
|
|
| 403 |
|
|
By -Meanwhile- - Posted on June 8, 2005 - 11:57:59 (#16836)
Current version when comment was posted: 0.0.2 |
|
|
"403 forbidden"...maybe a mirror would be nice/wise?
|
|
| 2 nicnac again |
|
|
By _3dEyes_ - Posted on March 28, 2005 - 06:02:24 (#16225)
Current version when comment was posted: 0.0.2 |
|
|
GOCR can recognition the information from screenshot, it has a mode optimized for this purpose. In the following version I shall make support of this feature.
|
|
| @nicnac |
|
|
By tb100 - Posted on March 28, 2005 - 05:45:53 (#16224)
Current version when comment was posted: 0.0.2 |
|
|
Screengrabs are not very good tests for OCR software - the resolution of screens is very much lower than scans (usually 3-6 times lower), and OCR programs are optimised to work best with scanned material. If you take a screenshot of an app with text 3-6 times bigger than usual it should give a more accurate idea of the accuracy of the recognition.
|
|
| 2 nicnac: |
|
|
By _3dEyes_ - Posted on March 27, 2005 - 15:23:48 (#16221)
Current version when comment was posted: 0.0.2 |
|
|
1. Try to increase DPI at scanning
2. I do not develop GOCR, it is more question to developers of this project (jocr.sf.net), i've made only the translator.
3. I now shall realize support of a database of images in the GOCRTranslator, it is possible will help to increase accuracy of recognition.
|
|
| Yeeeeeesss! |
|
|
By Kevin_Field - Posted on March 27, 2005 - 14:34:25 (#16220)
Current version when comment was posted: 0.0.2 |
|
|
This is so friggin' cool. Thanks for doing things the Be way.
Doesn't work perfectly on all forms, but still, awesome job nonetheless!
Kev
|
|
| Nice but a little bug on my pc |
|
|
By nicnac - Posted on March 27, 2005 - 13:55:23 (#16217)
Current version when comment was posted: 0.0.2 |
|
|
I save with Grabit the About GOCRTranslator
The result:
BeOS Im8ge_to_Text Tr8nsl8tor B8sed on OOCR Converts Im8ge wIth text to sever8I text tIle torm8
BeOS gr8phIcs 8pps usIng Tr8nsl8tIon KIt
All a are 8 ???
|
|
| Excellent implementation |
|
|
By stippi - Posted on March 27, 2005 - 13:33:29 (#16216)
Current version when comment was posted: 0.0.2 |
|
|
Very cool and BeOS-like implementation. The "simplest" solutions are always best!
Best regards,
-Stephan
|
|
| Superb |
|
|
By edmundf - Posted on March 25, 2005 - 04:17:33 (#16206)
Current version when comment was posted: 0.0.2 |
|
|
I was waiting for this for ages, now it is here and my first test
really went perfect.
I had a lot off trouble in .. yes windows, to OCR scanned pages
because the scanned tekst was rotated, in BeOS one can correct that
in a flash.
A big thanks for this, I am very very pleased with it.
|
|
| Good job ! |
|
|
By A-Quatro - Posted on March 25, 2005 - 03:57:34 (#16205)
Current version when comment was posted: 0.0.2 |
|
|
I was waiting for this kind of program, thanks !
|
|
|
|
|
