Recognition of ambiguous mathematical characters within mathematical expressions

Resolving ambiguity within mathematical symbols is essential for recognition of mathematical expressions. In this paper, we focus on the resolving ambiguities in mathematical symbols and propose a novel recognition technique that has been tested over large number of ambiguous mathematical symbols obtained from different categories including factoring formula, algebra identity, geometric progression, integral expressions, quadratic equations, area function, and geometric progression formulas for mathematical expressions. In this paper, we propose a novel feature extraction technique which helps in reducing the ambiguity within mathematical symbols. First, we have identified ambiguous isolated mathematical symbols from different categories of mathematical expressions. The features are extracted using feature extraction technique which includes Fourier descriptor and chain code. Two different classifiers SVM and K-NN are used for classification purpose. Fivefold cross validation technique has been used for computing recognition results. Overall recognition rate of 97.03 and 96.62% is achieved using SVM and K-NN classification respectively.