To create new models for Polish, you need to have transcribed text-lines (images + GT) either real or artificially generated. You can add polish characters to chars.py in ocrolib and then call them instead of default characters in ocropus-rtrain. Or, you can used -c option (use with caution, there is a little bug in the ocropus-rtrain code) to create codec from your GT data.
On Tuesday, November 4, 2014 2:24:27 PM UTC+1, [email protected] wrote: > > I'm intend to train the Ocropus 0.7 version for Polish language. If I > instead create new models, expand the default of Polish expressions? The > creation of new models of well recognized Polish learned words but wrong > about the words without Polish characters. > -- You received this message because you are subscribed to the Google Groups "ocropus" 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]. To view this discussion on the web visit https://groups.google.com/d/msgid/ocropus/ec29976e-4ee4-4b4a-8515-81cde4864698%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
