This is a learning project. The goal is to take images of handwritten equations and solve it. The equation can also be drawn by the user if wanted. The scope is it only processes quadratic equations. Hidden variable has to be x. For example
x^2-3x+2=0
The equation must contain x,y,0,1,2,3,4,5,6,7,8,9,+,- . Anything else wont work.
The project is run on python. Installing jupyter is higly recommended. There are 5 main files to see. Each part is given and detail is given in the code. MNIST dataset is used for training digits and x,y characters. + and - is trained from images downloaded from kaggle.
pip install jupyter
pip install keras
pip install tensorflow-gpu
pip install numpy
pip install opencv-python
pip install Pillow
pip install python-resize-image
pip install sympy
Run part4 among the codes. The rest is for training dataset. Choose to draw or select image files of equations. Set the model path accordingly test4.h5 is the latest. Other than that it would work.