If there originaly was a training data in the tessdata repository try using this one. Even in tesseract's official repository is stated that retraining is not likely to help you. You can try different scales of the image. Are you doing any image processing before giving the images to tesseract. You can familiarize yourself with how tesseract uses a dictionary, maybe there is the problem. Maybe you can process tesseract output - like calculate Levenstein distance from the output to some words from a dictionary.
Bear in mind that 100% recognition is impossible. I would say 80-90% is what you will probably end up with. 90%+ is extreme success in my opinion. Good luck. -- 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 https://groups.google.com/group/tesseract-ocr. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/99efcd9c-76ed-49de-ab08-12916b486df7%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
