FORMULA DETECTION MODULE Russian patent published in 2016 - IPC G06K9/34 

FIELD: data processing.

SUBSTANCE: invention relates to the field of detection and data processing. In the process of finding formulas formulas are located in the fixed format of the document by identifying the basic elements of the formula. Original formula element is a text element, which carries a statement that is part of the formula, such as a text string recorded the font used to display mathematical expressions and mathematical operators, symbols or keywords that are used in mathematical formulas. Create and expand a border around the original formula element to set the area of the formula. To eliminate the overlap with the surrounding normal text region the formula is divided into multiple regions, based on the formula of the vertical position and horizontal spacing between the elements of the formula. After their vertical ordering the horizontally overlapping portions are combined to recover the formula as a floating member.

EFFECT: technical result is to increase the rate of conversion of documents containing mathematical formulas, by automatically converting mathematical formulas elements in non-fixed format.

20 cl, 19 dwg