A Neural Network-Based Framework to Recognize the Handwritten Chemical Expression

A chemical reaction which written on the paper was termed as handwritten chemical expressions (HCE). These expressions are made up with organic structures, symbols, inorganic formulas, and operators. The recognition of HCE is the recognition of organic structures, symbols, inorganic formulas, and operators. In this paper, we had addressed the challenges faced to recognize the handwritten organic structure and symbols. Here in this paper, a neural network-based framework has proposed to recognize the HCE. This proposed framework accepts the scanned images of HCE as input. It has two major phases to recognize the HCE. In the first phase, it recognizes the handwritten benzene structures using a Radial Basis Function Neural Network (RBFNN) classifier and in second phase it will recognize the inorganic formulas and operators used in chemical expression using Support Vector Machine (SVM). This framework generates the chemical formulas of chemical expression as an output. The proposed framework has tested and verified with sample datasets which gives the promising results.