It is more realistic to model pixels in an images with some degree or type of correlation, because due to the point spread function and the intensity of a pixel may contain info from its neighboring pixels. Assuming that an image is a random field and each pixel is a random variable.

More Mohd Najib Ahmad's questions See All
Similar questions and discussions