Suppose, for example, that we intend to simulate the Weberull distribution of speckle noise to simulate the SAR image. I'm looking for detailed instructions from MATLAB to do this.
CODE: https://www.mathworks.com/matlabcentral/fileexchange/32041-automatic-robust-nl-means-denoising-filter-for-additive-and-multiplicative-noise?focused=5191118&tab=function
You need to scale it ...
https://earth.esa.int/documents/10174/669756/Speckle_Noise_Characterisation.pdf
In Matlab, you can use wblrnd or pdf function to generate noise following by the Weibull distribution. Look in the help file to get the detail.
You can treat the pixels intensity are out of the range [0, 255] by min/max values (0 or 255) or don't change those values.
If you use a vector signal, which I believe you do for simulating the SAR image, the Weibull noise or any other compound Gaussian distribution can not be generated by simple Matlab functions for a scalar random variables. The most robust method to simulate a compound Gaussian random vector is to use the spherical invariant random process (SIRP), especially if you want to impose a correlation matrix.
- Badges
- Science topic
- Similar topics
- Computer Science
- Images