For the transfer of mathematical knowledge from paper to electronic form, the reliable automatic analysis and understanding of mathematical texts is crucial. A robust system for this task needs to combine low level character recognition with higher level structural analysis of mathematical formulas. We present progress towards this goal by extending a database-driven optical character recognition system for mathematics with two high level analysis features. One extends and enhances the

more »

... al approach of projection profile cutting. The second aims at integrating the recognition process with graph grammar rewriting by giving support to the interactive construction and validation of grammar rules. Both approaches can be successfully employed to enhance the capabilities of our system to recognise and reconstruct compound mathematical expressions. The authors' work was supported by EPSRC grant EP/D036925/1. 1 A glyph is a single, connected, shape of pixels. Characters are often composed of more than one glyph, e.g. "j" contains two glyphs and "≡" contains three.