SudokuCV Solver

Installation

1. Clone the Repository:

   git clone https://github.com/andrewpols/SudokuCV.git

   Or download the .zip file of the project.

2. Set Up OpenCV and Tesseract Libraries:

   • All files associated with the TESSERACT library/dependencies should be included within the project structure.
     • If for some reason they are not, please refer to the following resources to download the requirements:
   • Download and install OpenCV for Java from here.
   • To use Tesseract OCR, download and install Tesseract and its Java wrapper (Tess4J) from here.
     • IF USING MACOS:
       • OpenCV is not directly supported for Mac. To download it, you must use some variation of a python environment to properly download OpenCV.
       • eg. Anaconda virtual environments OR using homebrew to download OpenCV
       • Provided is a link to a YouTube guide on how to do so.

3. Import Libraries into Your IDE:

   • Add the OpenCV and Tesseract libraries to your project's build path.

4. Configure Library Path:

   • If you're using a dynamic library (e.g., .dylib for MacOS, .dll for Windows), ensure that the library path is correctly set:

     -Djava.library.path=/path/to/libraries

   • Here is a tutorial in setting the .dll OpenCV library path for Windows

5. Build the Project:

   • Open the project in your preferred Java IDE (e.g., IntelliJ IDEA, Eclipse).
   • Ensure JDK is up-to-date (JDK 21).
   • Build the project to resolve dependencies and compile the source code.

6. Run the Program:

   • Ensure that the required libraries (OpenCV and Tesseract OCR) are properly configured and accessible in your project.
   • Run the main class (SudokuCV.java) to start the Sudoku Solver application.

Usage

• Project Goal:
  • The goal of the game is to fill a 9x9 grid with digits so that each column, each row, and each of the nine 3x3 subgrids contain all of the digits from 1 to 9.
  • A digit cannot be repeated in the same row, column, or 3x3 square.
• Playing the Game:
  • After running the program in SudokuCV.java, simply follow the on-screen instructions.
  • The user has two options to provide a Sudoku Puzzle:
    • Choose an already existing image of a Sudoku grid on their device
    • Allow a Video Capture that reads a Sudoku grid in real time
• Purpose:
  • After creating a Sudoku Solving algorithm, I found it was far too time-consuming to manually input the data of the grid manually.
  • Using computer vision automates this process, allowing quick, real-time solutions
• Tips For Use:
  • If using the video capture, try to ensure that you have the proper brightness/lighting for the camera to detect the grid. Inadequate lighting leads to rendering and processing issues.
    • Try performing the video capture in a well-lit room, and consider turning your computer's brightness up.

Tutorial: Image Examples

- Choose Image Pathway

- Live Video Pathway

Backend Images

• Extracted Cells