Exploring the Possibility of training Tesseract to recognize handwritten source code.

[Rajeev Kodippily]

Hi Folks,

I'm doing a study on handwritten source code recognition.

My goal is to convert handwritten source code ( obtained using stylus input, or it could be a written image if that is better suited for tesseract) to typeset.

I would like to explore any support tesseract may have for this particular use case. 

Would it be possible to train or fine tune tesseract for this use case ? I would like some initial feedback before I jump into trying this out.

Thanks!

Rajeev.