how to know noise type from standard deviation value

If i have a noisy image and I wrote a code to calculate its mean, variance.. etc. How can I determine the noise type and the appropriate filter type? ( I am working in App designer in matlab)

Answer 1

[Image Analyst]

You can't determine the noise type. There are lots of noise that could give rise to the same mean and stdev of the image. Anyway, why do your images have different types of noise instead of all having the same type of noise. There are a bazillion different denoising routines. Why don't you just offer the user a menu of common methods, like blurring, median filter, etc.? By the way, the mean and standard deviation of the image is not a good indicator of how much noise is in the image. Even a noise-free image will have values for mean and standard deviation.

[Zainab Riyadh]

Thank you of your response. It’s a course project where we are given different noisy images and we are supposed to write a code to determine the noise type for each image. I thought that each noise type has a range of possible mean or std values. So based on these values I can determine the noise type.

[DGM]

You presume that you can know the difference between the noise and the signal. Do you have noise-free reference images?

For example, both of these images have the same amount of the same type of noise.