Converting Handwritten Mathematical Equations to LaTeX Code

This project focuses on the conversion of handwritten mathematical equations into LaTeX code using computer vision and deep learning techniques. By training CNN models to recognize symbols, we developed a system capable of interpreting mathematical expressions and converting them into syntactically correct LaTeX.

Description:
Designed a system using image processing techniques and CNN models to classify handwritten symbols—including digits, Latin/Greek letters, and mathematical operators—and convert them into corresponding LaTeX code.

Tools & Technologies:
Convolutional Neural Networks (CNNs), Image Processing, Python

Outcome:
Demonstrated the capability of CNN models in accurately detecting and classifying mathematical symbols for LaTeX generation.

GitHub Repository:
Converting Handwritten Mathematical Equations to LaTeX Code

System Overview

the block diagram of the proposed method

The system architecture involves preprocessing handwritten images, segmenting symbols, classifying them using CNN models, and converting the sequence into LaTeX format. The pipeline effectively handles a wide range of mathematical expressions.

See Project Report for more technical details and evaluation.