HANDWRITTEN LINEAR EQUATION SOLVER

Abstract

Mathematics plays an important role in everybody’s life. We mainly depend on calculators available in our electronic gadgets to solve mathematical equations. Remembering the fact that almost all electronic gadgets are screen based where we can draw now-a-days, developing a system that recognizes and solves mathematical equations from handwriting is a potential area of research. This work develops an automatic linear equation solver, which solves the handwritten linear equations. Neural network approach is used for recognition of the digits, operators and alphabets. The system is successful in solving linear equations with 2 and 3 variables having finite, infinite and no solutions. The system gives a recognition accuracy of 96%. It is an interactive and user-friendly project where user can easily draw their equations rather than typing which consumes time and get their equation solved.