Sudoku Solver on python with OCR to read the sudoku from images
The OCR is a machine learning model trained on the MNIST dataset, procedure can be found in train_model.py
The sudoku that is extracted from the images is solved using a backtracking algorithm
If you wish to simply see the results then you can run solver.py and if you want to retrain the model the you can do it from train_model.py
P.S. Sample of the results of running solver.py in a terminal window are there in working.png