Preparation of a specific character-set traineddata

[Karol Wójcik]

Hi there, 

So far I've been using https://github.com/Shreeshrii/tessdata_shreetest/blob/master/digits_comma.traineddata. Generally speaking, with very good results, much better than when using eng-best or eng-fast from standard tesseract repo. But, unfortunately, recently I came across some unrecognized characters when ocr-ing my data sets and it seems it's blocking further development of my software.

I tried to fine tune it myself, but unfortunately the results got worse :( So I'm looking for somebody willing to create a specialized traineddata for me. It would require a few additional characters added along to digits_comma.traineddata. I would want to achieve the same accuracy as when using digits_comma.traineddata.  

I'd be more than happy to pay premium for such work.

Best Regards,

Karol

[Karol Wójcik]

Nobody? :(