how can a matlab code be written for identification of noise using statistical features like kurtosis and skewness?

noise identification in images using statistical features like kurtosis and skewness

Answer 1

[Image Analyst]

Try this: http://www.mathworks.com/matlabcentral/answers/8026-best-way-s-to-master-matlab

Skewness can indicate possible noise if the distribution of normal, noise-free pixels is known to be Gaussian. However that is often not the case and so skewness I don't think will be useful in general. There's even a less likely case that can be made for kurtosis. Even noise-free images could have a non-zero skewness and kurtosis, so what does that tell you?

[Pyla Sahitya]

I agree with u. But the thing is I need to do my project on noise identification in images using kurtosis and skewness. So I please request you to help me out in writing the code for the above in MATLAB.

[Image Analyst]

Attached is code for computing skewness and kurtosis. Please show me how you're finding noise using them, because, like I said, in general I don't think you can. Take an image (a real image like a snapshot) that's defined to be noise free, then compute skewness and kurtosis. Then add more noise to only some of the pixels. Then computer skewness and kurtosis on this noisy image. Now, find the pixels that had noise added to them. I'd like to see your findings. Or take some arbitrary image where you have no noise-free reference for. Now tell me if there's noise in it or not based on the skewness and kurtosis values.

[Pyla Sahitya]

yes sir definitely I will

Answer 2

[S C.Carl]

I have the same question. How to identify noise in an image ?

Dear Image Anaylst, could you please send me the application to identify noise pattern with skewness and kurtosis, if you have

Dear Pyla, could you please share the codes help me to solve this problem. I have the same problem with you now :(