How to train new fonts for an existing language

[Pedro Correia]

Hi there!

For the past couple days I've been using jTessBoxEditor in order to get the correspondent .box files of some images, for several fonts. Now I need to train these fonts along with the ones already trained on the english model, but I dont know how to make a single .traineddata out of all of this! How can I do it?

One thing that care to my mind was to train a model exclusively for these new fonts and them run tesseract with multiple languages (-l eng+NewModel). Would that yield results as good as it would if everything was a single model?

Thanks in advance!