I got it, the problem was if I don't specify the region of recognition. For
some reason it can't recognize it fully, but if make a border around that
text - all ok! I did the same in my code:
using (var engine = new
TesseractEngine(Application.StartupPath + @"\tessdata", "coc",
EngineMode.Default))
{
engine.SetVariable("tessedit_char_whitelist",
"0123456789");
engine.DefaultPageSegMode =
PageSegMode.SingleWord;
var result = engine.Process(b).GetText();
int.TryParse(result, out number);
}
if (number < 10)
{
using (var engine = new
TesseractEngine(Application.StartupPath + @"\tessdata", "coc",
EngineMode.Default))
{
engine.SetVariable("tessedit_char_whitelist", "0123456789");
var result = engine.Process(b, new Rect(0,
4, b.Width, b.Height - 8), PageSegMode.SingleWord).GetText();
int.TryParse(result, out number);
}
}
And now it is much more stable! Also, finally I trained data for special
font and accuracy of recognition is perfect now. Thanks to everyone!
суббота, 1 ноября 2014 г., 20:38:20 UTC+3 пользователь Quan Nguyen написал:
>
> The image is really small -- it needs 300 DPI.
>
> Nevertheless, VietOCR 4.0 beta, which uses Tesseract 3.03 RC, can pick it
> up without any problem.
>
> If you use the .NET version, be sure to scale the image first.
>
>
--
You received this message because you are subscribed to the Google Groups
"tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/tesseract-ocr.
To view this discussion on the web visit
https://groups.google.com/d/msgid/tesseract-ocr/42dd1515-cb2c-4ade-82e4-879b8caaa0aa%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
