Sudoku Solver with CNN and Backtracking Algorithm

This repository contains a Sudoku solver implemented using Convolutional Neural Networks (CNN) and the backtracking algorithm. The solver is built using the PyTorch framework. Sudoku is a popular number-placement puzzle that requires filling a 9x9 grid with digits while following certain rules.

The solver utilizes two main approaches:

1. Convolutional Neural Networks (CNN):

   • A CNN is trained to recognize digits in a 9x9 grid.
   • The trained model can extract the digits from an input Sudoku image or grid.
   • This approach helps in preprocessing and digit recognition.

2. Backtracking Algorithm:

   • After extracting the digits using the CNN, the solver employs the backtracking algorithm to solve the puzzle.
   • The backtracking algorithm is a recursive algorithm that explores all possible solutions and finds the correct one.
   • It efficiently handles the puzzle's constraints and reduces the solution space.

Repository Structure

The repository is organized as follows:

• model/: Includes the trained CNN model for digit recognition.
• notebooks: Includes the model notebook and solver notebook.
• test image: Include one test image for testing the solver.

Acknowledgments

• The CNN model architecture and training process were inspired by various online resources and tutorials.
• The backtracking algorithm implementation follows well-established techniques for solving Sudoku puzzles.

Contributing

Contributions to this repository are welcome! If you have any improvements, bug fixes, or new features, feel free to submit a pull request.