SolvingSudoku

Contributors:

• Aleksandra Nedić (SW 27/2019)
• Dunja Stojanov (SW 30/2019)

Problem: Recognising and solving sudoku puzzle from image using convolutional neural networks

Algorithms: CNN and backtracking

Data:

• http://www.ee.surrey.ac.uk/CVSSP/demos/chars74k/
• https://icosys.ch/sudoku-dataset

Validation:

• 80% of data used for training the model
• 20% of data used for validation
• Validation of the solved table using py-sudoku library

Python version: Python 3.7

To start up the project do the following steps:

1. Create virtual environment and activate it.

2. Run the pip command bellow to install all the required packages (Requirements are listed inside the requirements.txt file).

   pip install -r requirements.txt

3. Unpack digits.rar

4. Open project in PyCharm configure Pithon interpreter from your new virtual environment.

5. To solve Sudoku puzzle run script detect_digits.py. To train model run script training.py.