This paper proposes a system that understands mathematical expressions on binarized printed document images. The system first extracts a set of "primitives" from the image. Each of the extracted primitives is associated with a "bounding box" and its label. Using the attributes of the primitives, the system constructs an initial hierarchy. Construction of an initial hierarchy includes merging a group of primitives into a key word. Next, the system checks the validity of the hierarchy according to conventional mathematical syntax rules. If any syntax error is detected, the system makes attempts to correct the errors. The modification step includes reconfiguring the initial hierarchy, revisiting the original image for possible missing primitives, placing dummy primitives into missing spots in the hierarchy, and so on. The corrected hierarchical structure can be converted into the format for a particular publication system such as T/sub E/X.
Understanding mathematical expressions from document images
Proceedings of 3rd International Conference on Document Analysis and Recognition ; 2 ; 956-959 vol.2
01.01.1995
400926 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Understanding Mathematical Expressions from Document Images
British Library Conference Proceedings | 1995
|Automatic Understanding of Structures in Printed Mathematical Expressions
British Library Conference Proceedings | 2003
|On Machine Understanding of Online Handwritten Mathematical Expressions
British Library Conference Proceedings | 2003
|