Handwritten math symbols/digits recognition for educational syllabus

Abstract

live in the age of technologies. People are becoming in more need for consuming their time and in need for more helping software. The scope of this paper is recognizing mathematical symbols and digits and introduces a guideline for future ap- plications that addresses solving mathematical based problems. This paper introduces a project that is a platform for imple- menting some mathematical applications that will have many advantages in the education process. Students are o ered solu- tions to their mathematical problems with an easy way of cap- turing the handwritten problem, also teachers will nd it useful by checking their students' answers faster. And if other features is added to those application as conversing methods of interact- ing process, Some applications of those kind will help also little kids learn numbers and some simple mathematical operations as addition and multiplication with fun and joy of the speaking feature of the application. This project is proposing two system architectures. The rst one is a complete implemented Keras system and the second is Tensor ow architecture training sys- tem. Methodology is CNN. Dataset used contain handwritten digits and mathematical symbols.