Slightly harder psnr. http://en.wikipedia.org/wiki/Peak_signal-to-noise_ratio
Harder again... ssim http://en.wikipedia.org/wiki/Structural_similarity
There isn't any real solution to automating similarity as a human sees it. Humans are tricky.
He just loops over the whole image and does a pixel-by-pixel comparison, taking the difference between each of the R, G, and B channels and summing them all.
