Asked by Sajid Khan
on 21 May 2013

Hi everyone,

I am trying to generate training sequences of different noises but I am not confirmed that whether am doing it in a correct way or not.

Here is the code snippet for generating them,

%% for uniform image

% image = uint8(ones([512,512])*128);

A = -30;

B = 30;

matrix_uniform = uint8(A + (B-A)*rand(size(image)));

%for gaussian image

Mean_added = 0;

Variance_added=400;

a = ones(512)*128;

matrix_gaussian = uint8(a + Mean_added+sqrt(Variance_added).*randn(size(a)));

%% for impulse noise

image = ones(512)*128;

matrix_impulse = imnoise(uint8(image),'salt & pepper',.4);

%% for speckle noise

image = ones(512)*128;

matrix_speckle = imnoise(uint8(image),'speckle',.1);

Am I doing it in correct way.

Also how to find their kurtosis and skewness using Matlab, actually I want to get reference kurtosis and skewness to compare with the skewness and kurtosis of noise extracted from the image to check the performance of algorithm.

Answer by Image Analyst
on 5 Mar 2014

See my image moments demo, where I compute skewness and kurtosis.

(I know this is old, but perhaps someone else will like the demo.)

Opportunities for recent engineering grads.

## 0 Comments