RECOGNITIONOF LINES INTHE HISTORICAL HANDWRITTEN DOCUMENTS

Gippiev M.B, Zhukov A.V, Rogov A.A, Skabin A.V

208 KB

Transformation of handwritten or printed text images into text data has currently become one of the most actively developing areas in pattern recognition. There are many software tools which allow automation ofprinted text and text forms recognition, provided with a good digitizing of the source. This article deals with string extraction on handwritten documents, whichhave been digitized using a camera. String selection in printed and qualitatively digitizeddocuments is not a difficult problem, which can be solved using widespread methods. Comparisonof two new string recognition methods is discussed: method of nearest neighbor and method of bond graph construction. Algorithms, as well as the results of their work on stenographic documents andcriteria of evaluation of their performance are described in the article. Moreover, the article presents an algorithm of subscript and superscript character recognition.

Библиографическая ссылка

Гиппиев М.Б, Жуков А.В, Рогов А.А, Скабин А.В РАСПОЗНАВАНИЕ СТРОК В СТЕНОГРАФИЧЕСКИХ ДОКУМЕНТАХ // Научное обозрение. Физико-математические науки . 2020. № 1. С. 25-25;
URL: https://physics-mathematics.ru/en/article/view?id=22 (дата обращения: 24.06.2026).