Компьютерный фонд древнерусской скорописи |
|
|
|
Written by: Юрий Николаевич Филиппович, Анна Юрьевна Филиппович, Иван Анатольевич Зеленцов
|
Воскресенье, 07 Сентябрь 2014 |
The paper presents
the project to create an electronic corpus of Old Russian cursive. A model for
and algorithms of Old Russian cursive handwriting recognition
are being developed. Key aspects of recognition technology are a structural
approach, duplex recognition circuit (letter/word), recognition controlled by
hypotheses, vagueness in the description, the use of knowledge and
participation of the expert. A mockup of the recognition system is being
developed. The main components of the system are a tracer, word and letter
recognizers, a knowledge base containing information about letters and words
and a training module.
philippovich-zelentsov_elmanuscript2014 (607.4 kB)
|