I have experience with Python and OpenCV and I am proposing the following method for this project:
1. To generate random noise
Read original image
For 1 to 50:
For each pixel p(x,y) at location (x,y) in the original image (say red channel) :
Generate a random variable r from normal distribution with mean 0 and variance 50
n(x,y) = p(x,y)+r
Save noisyimage from array n
2. Once all the images are written
Initialise Sum(x,y) = 0 for each location (x,y)
For each noisyimage:
Read image
For each location (x,y) in the image :
Sum(x,y) = Sum(x,y) + n(x,y)
Diff(x,y) = Abs(Sum(x,y) - p(x,y))
That is, sum up the pixel value at each location (x,y) for all images and take the average. Take pixel wise difference between this average value and the original value for each location. Save the difference as an image and show that it is almost black.
Please feel free to write to me if you need any clarification. Thanks!