A Structural Analysis Approach for Online Handwritten Mathematical Expressions

Summary This paper proposes a structural analysis approach for mathematical expressions based on the Attribute String Grammar and the Baseline Tree Transformation approaches. The approach consists of geometrical feature extraction, parsing structure and expression analysis steps. The algorithm for structure parsing uses baselines, which are represented by geometrical features to recursively decompose the hierarchical levels. During a hierarchy decomposition of an expression, it predicts the relationships along the symbols, and then groups the symbols into a number of sub-expressions and a set of basic units. The basic units are added into the tree and the same parsing process is performed on these sub-expressions. Once all sub-expressions are parsed, expression analysis is carried out to translate the tree into any desired syntax (e.g. Latex, Lisp). The experiments are performed to test the efficiency of this proposed structural analysis approach.