在量化图像上基于RGB值分割图像

Question

How would I segment the background from an image that has already been quantized(n=5/6)?

Assuming, I have a clean background color that is only one color, I think filtering out everything by either the rgb value of the top left or top right pixel would work great.

Ideally, I would like to avoid using openCV

Sample:

Answer 1

Compare each pixel color to the pixel in the top-left corner. A pixel where all three layers of R, G, and B match is background:

mask = np.all(image == image[0, 0, :], axis=2)