This is project is aimed at recognizing, segementing, classifying and parsing handwritten math expressions.
*Phase-1 of the project i.e. classification of individual math symbols along with junk rejection has been completed as of 3/5/2018. *Phase-2 of the project i.e. Segmentation of an expression into individual symbols and passing on to the classifier engine has been completed as of 4/8/2018.
Phase-3 the final phase of the project has been completed. This phase built a parser that takes classified symbols and weaves them into a logically correct math expression.
For training, CROHME dataset has been used: https://www.isical.ac.in/~crohme/CROHME_data.html
Please find the README.txt file in project on instructions to extract features and run the models.
For more information on features and models refer the Phase-1-Classification.pdf file.
- = Old updates.