SUDOKU

Converts image of SUDOKU into digital form and finds a solution

• Using Python 3.6.8 and OpenCV 4.1.0

  How to install:

  • Install Anaconda

  • Clone this repository git clone https://github.com/horczech/sudoku.git

  • Install the virtual environment from enviroment.yml file located in the root directory of the repository using

    conda env create -f environment.yml

  • Activate the virtual enviroment

    conda activate sudoku_env

  How to use:

  • There are two scripts that can be used from command line.

    1 sudoku_solver.py that accepts SUDOKU image and returns image with solved SUDOKU and text version of solution.

    python3 sudoku_solver.py sudoku_imgs/web_cam/webcam_clean_1.jpg

    There is also optional parameter --config where you can specify parameters of the algorithm. The example config files can be found in /configs/ directory.

    python3 sudoku_solver.py sudoku_imgs/web_cam/webcam_clean_1.jpg --config configs/config_07

    2 camera_sudoku_solver.py This script has no input arguments it just finds and solves soduku from the camera stream. By pressing "F" key on the keyboard it will freeze the surrent frame from camera and by pressing "R" key it will return to the camera stream. By pressing "P" key it will wait till it finds valid solution of the SUDOKU and prints image with the solution and text version of solution. Run it using

    python3 camera_sudoku_solver.py

  Sources:

  • http://www.aishack.in/tutorials/sudoku-grabber-opencv-plot/
  • https://hackernoon.com/sudoku-solver-w-golang-opencv-3-2-3972ed3baae2
  • https://ieeexplore.ieee.org/document/7007986?tp=&arnumber=7007986