short intro to this topic :
abbyy allows generation of new language models based upon existing
or newly formulated ones.
I am trying to build up a language that contains specifically almost all
symbols which are needed to OCR a book(based on Latin some greek etc.).
Nonetheless when i look at the recognized pages the
symbols, which i specified, like Ⓐ or ①②③ are not used.
Is this due to missing fonts.
Has the training of the user pattern anything to do with this.
The book is quite large and it makes no sense to do this manually.
ideas any1
