Color-Based Segmentation Using K-Means Clustering

This question already has an answer here:

  • How to detect colors under different illumination conditions

    1 answer

Hi i found this tutorial about segmenting colors from images in

http://www.mathworks.com/help/images/examples/color-based-segmentation-using-k-means-clustering.html

There is a part there wherein the colors has been segmented into 3 part

specifically this part of the code

    for k = 1:nColors
     color = he;
     color(rgb_label ~= k) = 0;
     segmented_images{k} = color;
    end

Now the output is partitioned into 3 separate colors

IE.

image 1 contains only blue

image 2 contains only yellow

image 3 contains only brown

now what i am asking is how can i just get the brown partition? In my example the position of the brown color is 3 but sometimes when i partition other images, The position of the brown color becomes 2. How can i determine what color goes to which partition when using the LAB colorbased segmentation?

Thanks in advance 🙂

You can use the mean of each group to decide and compare it to a standard mean that you define. That way, you can algorithmically define which classified group is close to your “brown” group and use the brown color for it.