Skip to content

A simple Rubiks Cube solver that gives the optimal way to solve the cube.

License

Notifications You must be signed in to change notification settings

KKVANONYMOUS/Rubiks-Cube

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Rubiks-Cube

Description

The Rubik’s Cube is an infamous puzzle designed in 1974 by a Hungarian professor, known as Ernő Rubik. The puzzle consists of a large cube being split into a 3x3 orientation of smaller cubes, with each side being able to rotate 360 degrees. Each side is assigned a colour, and the cube is considered solved when each side is made up of just one colour.
Every time you rotate a side, it breaks the colours apart, scrambling the puzzle. Once the cube is fully scrambled the aim is to try and piece all the colours together until it is back in its solved form.

There are various methods to solve Rubik's Cube but here we will make use of CFOP Algorithm to solve the cube in a very optimized manner.

Sounds interesting,right?

Let's dig deep into the various aspects involved in solving the cube

Theory

How the program takes Cube Input?

  • Images of all the cube faces are given as input to CubeInput.py file (See Giving Input to know how to take cube images) where it reads all the images one by one and stores the image data in the form of numpy array. To crop the unnecessary part of image, we have used selectROI() function.
  • The croped image is then resized and pass to face_color() function where the different color blocks are masked out using their hsv values and inRange() function of opencv. This function returns a black-white image of each block where the particular color is white and rest is all black.
  • This image is passed to mask() function where we have find contours and then iterated over each contour and rejected small contours on the basis of area.We have initiliased a 2-d zero array in which we store the color states of each face. Using some maths and using centres of each block ,we have filled the 2-d array.
  • The blocks are given color according to the manner specified in Notations section.

Structures used to store the cube state

  • arr is a 2D vector which consists of all 6 colours and its index represents its neighbours. i.e. in arr[i][j], i denotes the colour code and j represent the color code of its neighbors.

  • matrixes[6][9] is a vector of struct faceMatrix consisting of two vectors v1 and v2 where v1 stores its current position of that block and v2 stores its current position according to the centre, (for example: matrixes[1][1] looks for current position of 2 sided red-yellow block and v2 represent in which center red color is faced).

Algorithms involved in solving the cube

As discussed above CFOP Algorithm is used for solving the cube. The various steps involved in this algorithm are as follows:

  • Making White Cross :- For making white makeWhiteCross() function is used.There are total 4!=24 (solving in any sequence of red, green, blue and orange) ways to make white cross. We check for all the possible ways and then the most optimal way (that takes minimum number of steps) is applied to make the white cross.

  • F2L :- The next step is to make the first two layer (i.e- middle layer of red, green, blue and orange and all four white corner pieces) which is implemented using the implementF2L() function based on the F2L Algo. Similar to the previous step, here again we loop through all the possible ways and implement the most optimal one.

  • Making Yellow Cross :- First we check if the yellow cross is already made or not by using the checkYellowCross() function. If it's not, we call the createYellowCross() function to make the yellow cross by looking at all possible cases.

  • OLL :- When the first two layers (F2L) are solved we need to orient the last layer (OLL) so the upper face of the Rubik's Cube is all yellow. We don't care if the side colors don't match, we are going to permute the last layer (PLL) in the next step. The function oll() implements the above algorithm by first counting the number of yellow corners in the current cube state and then applying the corresponding algorithm for orienting the layer.

  • PLL :- In this final step we make use of the setCorners function to set the corners of the last layer and then call setEdges to finally set the edges.

Hurray our cube is solved! :)

Usage

Notations

Every rotation of the cube is represented by a unique color code lying between 0-11 (both inclusive) which are mentioned below:

          COLOR CODES
 -----------------------------------
 Color  |  Clockwise  | Anticlockwise
 ------ |-------------|-------------
 white  |     0       |     6
 red    |     1       |     7
 green  |     2       |     8
 blue   |     3       |     9
 orange |     4       |     10
 yellow |     5       |     11

For example, color code 5 means that the yellow face needs to be rotated only once.
Similar logic can be applied for other codes as well.

Installation

  • Clone the repository
$ git clone https://github.com/KKVANONYMOUS/Rubiks-Cube.git

or download zip file

  • Change your current directory to the repo's root.
$ cd Rubiks-Cube

Giving Input

  • Click the images of all the 6 faces in the following specified pattern:
    Consider we are first taking the image of cube with red centre,
    Now keep the current face towards you, i.e. along the x-axis, faces with white and yellow centre along the z-axis with the white centre one facing the negative z-axis and finally the blue and green centre faces along the y-axis with green centre at your right side (i.e. negative y-axis).

    Follow similar kind of pattern for all the remaining faces.

  • Name all the images in the format- 1.jpeg,2.jpeg and so on and save it in the same directory in which CubeInput.py file is present.

  • Run the CubeInput.py file using:

    $ python3 CubeInput.py
  • This python file will return the states of all the faces in color.txt file as shown in Sample_Input_For_CFOP.txt. Copy paste the content of this file and give this as an input to CFOF.cpp file.

Solving the cube

To solve the cube, run the CFOF.cpp file by using

g++ CFOP.cpp -o CFOP

and give the content of color.txt file obtained in Giving Input section as an input to this program.

After the completion of each step involved in CFOP Algortihm the program will output the number of steps and their respective orders used to implement it.

For example, the following snippet shows the steps involved in making the white cross

Total Steps for WHITE CROSS: 7 //No of Steps
10 2 4 3 4 3 3 // Order of steps

Please refer to Notations for cube notations.

Demo Input

The repo contains a demo directory with images folder for sample input for the CubeInput.py program which will generate the corresponding face matrix, to be given as input to CFOP.cpp program. Or else the user can directly give the face matrix present in Sample_Input_For_CFOP.txt file as input to CFOP.cpp file.

For more info regarding giving input to the program you can refer to Giving Input section.

Shuffle function

This function gives additional option to shuffle the cube in order to test the code for different cases. Just give the cube rotation input in the form of numbers ranging from 0-11 as decribed in Notations and press 12 when you are done with the shuffiling.

Team

About

A simple Rubiks Cube solver that gives the optimal way to solve the cube.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •