Apply Singular Value Decomposition on input image

Question

I want to apply SVD to an input image and do a feature extraction. This is my code snippet:

[row, col] = size (A);
for i = 1 : row
    for j = 1 : col
        [U,S,V] = svd(A(i,j));
    end
end

Here A is the input image whose row = 347275 and col = 64. After executing the code I get U = 1, V = 1 and S = 1.4981. I know there is some mistake in my code. Can anyone suggest the right thing?