Модель системы распознавания старопечатных кириллических текстов с помощью лингвистической базы данных
Автор(ы): Денис Рашидович Касимов, Александр Валерьевич Кучуганов, Максим Николаевич Мокроусов, Павел Петрович Осколков   
07.09.2014 г.
In this paper a technique and an experimental system of optical character recognition of early printed Cyrillic texts are proposed, which are characterized by building an attributed graph of an image containing fuzzy attributes of the outline and skeleton of symbols generated by detection of letters on the basis of description logic as well as by the method of fuzzy recognition of words by building a tree of correspondences to a linguistic database. icon kasimov-kuchuganov-mokrousov-oskolkov_elmanuscript2014 (1.38 MB)