Clustering the image using k means

I have detected the face and have extracted features for face such as mean ,variance ,standard deviation,

I have applied k means directly to image and have clustered,by converting to HSV,now i have to give comparison result

by applying kmeans on features extracted values on the image ,please tell how to perform this

Answer 1

[Image Analyst]

scatter() is often used to compare results by showing the clusters. Have you tried using scatter to visualize your cluster results?

[nkumar]

k means on image
I = imread('');
I = im2double(I);
HSV = rgb2hsv(I);
H = HSV(:,:,1); H = H(:);
S = HSV(:,:,2); S = S(:);
V = HSV(:,:,3); V = V(:);
idx = kmeans([H S V], 4);
imshow(ind2rgb(reshape(idx, size(I,1), size(I, 2)), [0 0 1; 0 0.8 0]))

k means on values
I = imread('');
I = im2double(I);

m=mean(I(:));
va=va(I(:));
r=[m va]
idx = kmeans( );

how to apply k means for r to display image like above

[Image Analyst]

You said "I have applied k means" but it appears that you have not. I don't have the Statistics Toolbox so I can't try your code or develop any using kmeans(). All I can suggest is this Mathworks example: http://www.mathworks.com/products/demos/image/color_seg_k/ipexhistology.html

Answer 2

[Thorsten]

In your example, use

kmeans(r, number_of_clusters)

[nkumar]

I = imread(''); I = im2double(I); m=mean(I(:)); va=var(I(:)); r=[m va] idx = kmeans(r,2 );

in idx i get only 2 values,how that cluteres image can be image can be diplayed

[Image Analyst]

You asked for 2 and got 2. Did you run the Mathworks Example to see how they can be displayed?

[nkumar]

yes but that consists of many rows and columns ,but i have only 1row and 1 column

[nkumar]

i have a image below

http://imgur.com/qP3J654

i have ectracted features

features=[m v s] for an image it is dispaled as

features=[10 12 0.9]

now as in image how to perform Image clustering

please assist

[Image Analyst]

I don't have the stats toolbox so I can't try anything myself. Why don't you call tech support?