Noise

Noise may appear in an image in many ways: during image capture, transmission, storage, as well as during image copying, scanning or displaying. Some algorithms in computer graphics generate specific noise, too.

There are many different types of noise. Here, I shortly present two common-known types of noise in image processing : Gaussian noise and Impulse noise, also known as salt and pepper noise, or combination both of them. Salt and pepper noise scattered throughout the image in such a way pixels cause white and black points appears in digital grayscale images, which chaotically scattered along image area. It is defined by noise density. While Gaussian noise, also know as random is expressed in terms of its mean and variance values.

Here, I present very simplified formula for noise image, where $y(i, j)$ - output image, $x(i, j)$ - an uncorrupted image, $n(i,j)$ - noise model.

Below, I present a few images with a noise distribution in the following way :

where, $m(i,j)$ - Gaussian noise $N(0,\delta^2)$ and $p(i,j)$ is impulse noise).

Now, I present the image 'secret garden' with 'salt and pepper' for P = 0.02, 005 and 0.1, respectively.

But, here I present the images with 'gaussian' noise for $delta$ = 0.02, 005 and 0,1, respectively.