Title :
Structural Analysis of Printed Mathematical Expressions Based on Combined Strategy
Author :
Ha, Ming-Hu ; Tian, Xue-dong ; Li, Na
Author_Institution :
Coll. of Math. & Comput., Hebei Univ., Baoding
Abstract :
Recognizing mathematical expressions from document images is a key problem in the conversion of scientific document into electronic form. It is also a difficult part in the development of recognition technology. In this paper, we present an efficient method of parsing printed mathematics notation, which is based on the combined strategy of baseline and minimum spanning tree method. The syntactic and semantic knowledge is summarized so that a logical arrangement of the mathematical expressions can be obtained automatically. Experiments have been carried out for many types of expressions in printed documents and our method has shown favorable results
Keywords :
document image processing; grammars; optical character recognition; trees (mathematics); baseline method; document image processing; electronic document; mathematical expression recognition; minimum spanning tree method; printed document expression; printed mathematical expression structural analysis; printed mathematics notation parsing; scientific document conversion; Application software; Character recognition; Cybernetics; Educational institutions; Image analysis; Image converters; Image recognition; Image reconstruction; Machine learning; Mathematics; Natural languages; Optical character recognition software; Pattern recognition; Mathematical expression recognition; baseline feature; minimum spanning tree; structural analysis; syntactic and semantic analysis;
Conference_Titel :
Machine Learning and Cybernetics, 2006 International Conference on
Conference_Location :
Dalian, China
Print_ISBN :
1-4244-0061-9
DOI :
10.1109/ICMLC.2006.258474
