0

Hi im using https://github.com/charlesw/tesseract dotnetwrapper for tesseract3 it works but translating is very bad, My code is :

            TesseractEngine ocr = new TesseractEngine(null, "eng", EngineMode.Default);
            Pix pix = PixConverter.ToPix(bitmap);
            Page page = ocr.Process(pix);
            string result = page.GetText();
            MessageBox.Show(result);

and ;

Image img = (Image)bitmap;

Clipboard.SetImage(img);

above image(its not image i know but i dont got 10 rep to post image so consider it image) is translates as ;

Image in; = (Image)biI1Iip;

(lipbuard.SetIlIige{illg)5

i want it translate full so where is my problem,when i use same image in gttext program it translates full,so where is my problem.Thanks everyone who will try to give solutions.

1 Answers1

0

I assume that you don't use any image processing before passing the image to tesseract.

Here are some tips you can use to improve the quality of the result:

  • Fix DPI (if needed) 300 DPI is minimum
  • Fix text size (e.g. 12 pt should be ok)
  • Try to fix text lines (deskew and dewarp text)
  • Try to Fix illumination of image (e.g. no dark part of image)
  • Binarize and de-noise image

Source: image processing to improve tesseract OCR accuracy

Community
  • 1
  • 1
George Chondrompilas
  • 3,167
  • 27
  • 35