| Recognition of mathematical notation has been investigated for over thirty years. It plays an important role in transcribing documents in scientific and engineering disciplines from images into more abstract electronic forms. These allow the mathematical expression to be evaluated or manipulated by computer software.; Recognition of mathematical notation consists of symbol recognition and symbol arrangement analysis. In this thesis, we focus on the latter. In particular, we focus on the uncertainty occurring in the spatial structure between symbols. We investigate extensions to an existing deterministic mathematical notation recognition system DRACULAE, which is grammar-based and has potentials to extend to a Fuzzy Logic approach.; We develop a Fuzzy Logic approach to recognition of mathematical notation by applying Fuzzy Logic to the symbol arrangement analysis. By defining confidence values for spatial relationships and baseline structure trees, we produce a list of reasonable interpretations and allow users to select the desired interpretation of their expression. |
