Let's say I have a sample of pixel values of a photo, so the values themselves are in some recorded intensity. However, there is an error associated with each pixel which is by default set as 20% by the instrument that recorded the photo. I'm interested in finding the uncertainty of the mean of the sample of pixels. We can assume that there are $n$ pixels.
I know that usually one can consider the standard deviation ($std$) of the sample and find the uncertainty of the sample's mean by taking the $std / \sqrt n$. However I'm not sure what to do here, as each pixel itself has an error associated. Any answers will be much appreciated. Thanks a lot!