An approach for recognition and interpretation of mathematical expressionsin printed document

In this paper, we propose an approach for understanding Mathematical Expres sions (MEs) in a printed document. The system is divided into three main co mponents: (i) detection of MEs in a document; (ii) recognition of the symbo ls present in each ME; and (iii) arrangement of the recognised symbols. The MEs printed in separate lines are detected without any character recogniti on whereas the embedded expressions (mixed with normal text) are detected b y recognising the mathematical symbols in text. Some structural features of the MEs are used for both cases. The mathematical symbols are grouped into two classes for convenience. At first, the frequently occurring symbols ar e recognised by a stroke-feature analysis technique. Recognition of less fr equent symbols involves a hybrid of feature based and template-based techni que. The bounding-box coordinates and the size information of the symbols h elp to determine the spatial relationships among the symbols. A set of pred efined rules is used to form the meaningful symbol groups so that a logical arrangement of the mathematical expression can be obtained. Experiments co nducted using this approach on a large number of documents show high accura cy.