Recognition of handwritten mathematical expressions using machine learning

Authors

  • Sagar Shinde, Mukil Alagirisamy

Abstract

Handwritten digits, symbols and handwritten mathematical expressions recognition is a
demanding or inspiring mission owing to variation in writing style its , overwriting in symbols, characters ,
digits also complex semantics and two dimensional structure. It is an area of research from last three
decades, but not a single method fulfills all the requirement viz. execution time and identification rate. The
database collected or generated from various writers considering the style of writing, stroke movements,
overwriting and all that also the statistical and complex features have been detached and neural network
classifier has been used for the recognition. The automatic recognition of math symbol and expressions are
necessary to recognize math symbols, expression, digits, characters in students answer sheet, Postal address,
handwritten notes etc. The success rate of the system can be determined with throughput, efficiency and
identification rate.

Published

2020-05-30

Issue

Section

Articles