color image quantization between ABC algorithm and Fuzzy c-means fcm

I have a hybrid approach that combines the algorithm of the artificial bees and K-Means algorithm.

The problem is that I want to replace k-Means Algorithm and replace it with fuzzy c-means Algorithm(fcm in matlab) , knowing that the code has been implemented to quantized the color image.

This particular code is the k-means function. I want to replace it with (fcm) without changing the ABC code.

function [pos, cost,IDX] = fns1(X,pos,nColor,MaxIter)

for i=1:size(pos,1)

% perfrom kmeans

startCentroidTemp(1:nColor,1:3) = pos(i,:,:); % initial centroid should be k by p

[idx,C,sumd] = kmeans(X, nColor, 'MaxIter', MaxIter, 'Display','off', 'Start', startCentroidTemp);

cost(i) = 1/nColor * sum(sumd); % average distance

pos(i,:,:) = C;

IDX(i,:) = idx;

end

end

PARAMETERS

size of image=512x512 RGB

population=10=N

pos=10x4x3 double

cost=1x10 double

IDX=10x262144 double

X=262144x3 double

NColor=4

MaxIter=2

idx=?

C=4X3 double

Sumd=?

Use matlab 2019a

[mohammed altalibi]

hi