Understanding mathematical expressions in a printed document

摘要

A system for understanding mathematical expressions is presented.After separating all symbols in an input mathematical expression, weutilize thirteen features to represent each symbol. In order to reducethe computational time, a coarse classification algorithm is used toreduce the number of candidates. Then for each input symbol, thecharacter with the highest similarity is selected as the candidatesymbol. Since some of the symbols in an arithmetical expression maytouch each other, a dynamic programming algorithm is adopted to identifycorrect characters from connected symbols. In the expression formationstage, a procedure-oriented method is proposed for translating therecognized symbols appearing in a 2D space into a 1D character string.The authors have used 105 mathematical expressions as training data and50 expressions as testing data. The experimental results havedemonstrated the feasibility of the understanding system