There is no way to add one font to traineddata, it has to be done on all
fonts at one time
The box tiff pairs are not provided by Google developers who provided the
traineddata (some fonts are proprietary).
You can attempt to recreate the training using the source files from
langdata repository.
I have tested the traineddata concatination using "+" operator. But my
question is whether there exist a way to train one more font to existing
trained data ? or does it need entire box-tif pairs at each time ? if
required, from where can I get this box-tif training data set of current
2 matches
Mail list logo
