Extract noise and apply it at other data

Hello, I have a data with noise. Is it possible to extract the noise from the data and to get the property/character of the noise by using Matlab? Because I want to use and apply the same noise at other pure data(without noise).

Answer 1

[Walter Roberson]

No, there is no way to automatically tell the difference between noise and information.

When RAM was invented (and for many years after), they used to have problems with sudden random bit errors in the memory. They investigated and found it to be due to cosmic rays. Physicists got more interested in studying cosmic rays and started researching building detectors. Until someone realized that one of the best detectors around was RAM-style chips. So what used to be useless noise in a system turned out to very interesting information...

Answer 2

[Image Analyst]

I think only if you make some assumptions about either your signal, your noise, or both. Like your signal has low spatial frequencies or something like that.

Answer 3

[Arsalan]

If you are able to produce the original data which has not passed through any system, attain its spectral proporties, followed by finding an approximate transfer function for the system which the original data is to be passed through. and finally attaining the spectral proporties of the data you have.

By using these information and their comparisons you'll be able to get some idea about the noise. What type of system are you using?