I want to make a program that solves the Rubik's cube.
Basically, I want the easiest way to get an output like this:
G G Y
W W Y
B G G
From an image like this:
.
I want to make a program that solves the Rubik's cube.
Basically, I want the easiest way to get an output like this:
G G Y
W W Y
B G G
From an image like this:
.
Using only colors for the recognition is a very problematic because the colors are very sensitive to change in the light of the scene. Therefor if you insist on using only color i would do the following:
For a new image first do white balancing second use the sampled values to segment the image according to each color.
My suggestion is use also the shape of the rectangles and not only the color for the segmentation.