Read in an image and scale the intensity values from 0 255. Add noise to the image with a mean and standard deviation of 20 using the randn command. Take the average of N noisy versions of the original (using independent noise samples (randn commands)!) for values of N = 0, 5, and 10 and compare the result to the original image using the mean-squared error (MSE). Does the MSE decrease with N or the sqrt(N)? The MSE is a scalar and is calculated as (original image averaged image)2