Automatic Classification of Spatial Relationships among Mathematical Symbols Using Geometric Features

Abstract

SUMMARYMachine recognition of mathematical expressions on printed documents is not trivial even when all the individual characters and symbols in an expression can be recognized correctly. In this paper, an automatic classification method of spatial relationships between the adjacent symbols in a pair is presented. This classification is important to realize an accurate structure analysis module of math OCR. Experimental results on very large databases showed that this classification worked well with an accuracy of 99.525% by using distribution maps which are defined by two geometric features, relative size and relative position, with careful treatment on document-dependent characteristics.

Mathematical Expression Recognition based on Probabilistic Grammars

“…Also, it might be interesting to consider adding a rejection class, to detect when two subexpressions are unrelated. Finally, these features can be applied to printed expressions and compared with earlier work [Aly et al, 2009].…”

“…• Spatial relationship classification has high error rates for distinguishing between right, superscript and subscript relations. It could be interesting to train di↵erent classifiers depending on symbol categories as it improves classification results between symbols [Aly et al, 2009].…”

“…The spatial relations can be defined only between pairs of math symbols [Aly et al, 2009] or between symbols and subexpressions [Simistira et al, 2014]. Determining relationships only between symbols is less complex, but some structures could require special treatment because the relations may involve sets of symbols.…”