A Sudoku puzzle solver that reads the puzzle from images. It works with an image file (.png or .jpeg) or Web Cam feed. The solution will be printed on the input image/ live feed itself.
Run the main.py file to open up the graphical interface.
The grid_analyzing/vision_solver.py file contains the endpoints of the program (if the GUI is not going to be used).
Note: Make sure the input image is not too small in size.
The digit recognition is done through a neural network, and data set generation script and a training script is also included in this repository.
