Image Averaging

Let’s consider a noisy image g(x,y) formed by the addition of noise n(x,y) to an original image f(x,y);

g(x,y) = f(x,y) + n(x,y)

Learn more on Noise Types!

The assumption is that noise at each pixel is uncorrelated and has zero average value “white noise”. It is a simple statistical problem to show that if an image g(x,y) is formed by averaging M different noisy images, then expected value of g approaches original image f:



next up previous