Analysis And Reconstruction Of Special Structure In Printed Mathematical Expressions

With the development of science and technology, people's demand on information transmission and information retrieve technology becomes higher and higher. Traditional paper document is indispensable, but it is no longer suitable for the rapid spread of information. Though the technology of OCR achieves the conversion from paper document to electronic document and promotes the development of information technology, it is still unripe in recognizing the mathematical expression, which is an important part of scientific literature. Therefore, the recognition of mathematical expression has attracted the attention of many researchers.The recognition process of printed mathematical expression mainly consists of four steps:expression extraction, segmentation and recognition, structural analysis and reconstruction, of which structural analysis is the key step, and reconstruction is another important step. Through researching and analyzing special structures of mathematical expression, that is, heap symbol, headgear symbol and matrix, and so on, this paper summarizes features and syntactic information of the special structures. On this basis, this paper frames corresponding structural analysis algorithm, and puts forth the recursive reconstruction algorithm for the special structures of mathematical expression by using the depth traversal strategy. The experiments on analyzing and reconstructing the special structures show the efficiency of this algorithm.