Finding Similarity measure between different images?

Question

problem Statement :

I have to make a table consists of ( i,j,s(i,j)) where i and j are Keyframes or images and S(i,j) is similarity measure value between those images.

how to find similarity value between two images?

Can anyone please tell me how to find this Using sum of squared distances of pixels using Matlab ?

My problem contains series of images , say N and i need to calculate total of N(N-1) similarity values.

please give me some incite on how to do programming for the same problem.

Thanks in Advance Krsna

Answer 1

You can find the mean squared error (MSE) of two images as follows (Edited):

e = abs(Y - X);
MSE = sum(e(:).^2)/prod(size(e));

The peak signal-to-noise ratio is often used to measure the distance between two images (often between a noisy image and the original image):

PSNR = 10*log10(255^2/MSE);