Rubik's Cube Solver

This project, developed by Stefano Bonetto, serves as a submission for the "Signal, Image and Video" course at University of Trento, part of the Artificial Intelligence Systems master's program.

The aim is to provide a straightforward solution to solving a Rubik's cube, leveraging a color detection method. This approach doesn't use any machine learning algorithms for the color detection, instead relying solely on image processing techniques to analyze the values of each processed pixel.

This project is divided into two primary components. The first involves recognizing the cube and detecting the position and colors of each cell to determine the cube's current state.

A key assumption is that the solver directs the user to follow a specific sequence for accurately scanning the cube. This premise forms the basis from which I developed my solution.

The scan of the cube follow this order:

• face with the yellow central cell
• face with the blue central cell
• face with the red central cell
• face with the green central cell
• face with the orange central cell
• face with the white central cell

We obviously need to pay attention to the orientation of the face.

The cube-solving component of the project utilizes the kociemba algorithm.

This algorithm generates a sequence of moves, adhering to this notation. To make these moves more user-friendly, I've integrated corresponding arrows into the webcam feed for each move, providing intuitive guidance for the user on how to proceed.

Project Layout

└──faces	                # ROIs
├──pngs	                        # stuff for GUI  
├──rubik_venv
├──scan                         # original images from webcam
├──squares                      # each square of each face 
|      ├──1
|      |  ├──square_1.png
|      |  ├──square_2.png
|      |  ├──---.png
|      |  └──square_9.png
|      ├──2
|      ├──3
|      ├──4
|      ├──5
|      └──6
├──README.md
├──REPORT_Rubik_Cube.pdf
├──clear.py                    # this script helps me to clear all folders
├──main.py
└──requirements.txt

Requirements

To deploy this project is required a python environment (> 3.6). Is also suggested to build a virtual environment:

python3 -m venv rubik_venv

to activate it:

source rubik_venv/bin/activate

and install all the requirements for this project:

pip3 install -r requirements.txt

Run

To launch the application, simply enter the following command:

python3 main.py

Other Resources

I prepare a complete report of my work, you can find it here.

You can find my presentation of the project here

Contact me

Stefano Bonetto: [email protected]