Sudoku Solver

Introduction

I am going to attempt to write a sudoku solver that integrates with computer vision (CV).
I believe this is the Hello, World equivalent in the realm of CV.

When completed, I should be able to point my phone’s camera at a sudoku puzzle and it will return an augmented stream with the solution overlayed.

I’ll post all the code on my github

Overview

I’m breaking this down into 7 parts. Links will be embedded below as those articles are written.

1. Sudoku Solver / Puzzle Generator
   1. Faster performance (Part 1b)
   2. Final performance (Part 1c)
2. CV tool for digit recognition
3. CV tool for board recognition
4. Image augmentation to overlay solution
5. Extend support from images to video stream
6. Wrap it up in an API
7. System Hardening