Algorithm 1: Online K-Means clustering algorithm for color-based image segmentation
1:	Initialize the ‘k’ number of centroids, $C_1$, $C_2$, $C_3....C_k$ with random values.
2:	Initialize the counts n1, n2, n3 .... nk to zero.
3:	while ‘pixel stream continues’ do
4:	p ← RGB2YCbCr(p)
5:	Match the input pixel, ‘p’, to a single centroid $C_i$ by minimizing the distance $\Vert p-C_i{\Vert }^2$
6:	Increment ni
7:	Update the matching centroid, $C_i$, using moving average
8:	$\widehat{C_i}$ ← $C_i$ + (1/ ni)(p − $C_i$)
9:	Classify the input pixel, ‘p’, as ‘i’.
10:	end