Improve recognition with multiple font sizes

[Brad]

When I OCR the attached image, I do not get any valid results.

If I break the image into two separate images (one for each word) recognition works.

If I shrink the height of the second word to be closer to the height of the first word, recognition works.

I was curious if anyone had some insight into a parameter I could set that would remedy this issue.  The only alternative I see would be to split this into multiple segments and run OCR twice.

[Dmitri Silaev]

Such params are not known to me. But if they were I'm pretty sure that would be a quite unreliable solution. In my opinion just stick with the solution you found yourself - split into fragments.

Best regards,
Dmitri Silaev
www.CustomOCR.com

--
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 tesseract-oc...@googlegroups.com.
To post to this group, send email to tesser...@googlegroups.com.
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/202bd49e-0756-4c0f-a977-cd8ded2a5e4c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.