How to add impulsive and gamma noise to an image without using imnoise function in matlab? -

i wondering if give me advice on how add impulsive , gamma noise image separately? it's easy imnoise function of matlab, i'm not allowed use imnoise, our ta said can use rand function.

i came across code seems doesn't work should(for impulse noise):

    noisyimage=originalimage+255*rand(size(originalimage));  

there's few issues line of code:

• if image rgb (e.g., loaded colour jpeg or something), you'll have add 3 layers of noise; e.g., use

  rand([size(originalimage) 3]) 

• 255*rand() generates double-valued numbers, whereas image of type uint8 or (check class(originalimage)). fix, use randi instance:

  noisyimage = randi(255, [size(originalimage) 3], class(originalimage)); 

• you add noise of maximum magnitude 255 all pixels. might overflow many of pixels (that is, assigned values higher 255). avoid, use

  noisyimage = min(255, originalimage + randi(...) ); 

• the noise direction positive. true noise brings down values of pixels. so, use

  noisyimage = max(0, min(255, originalimage + randi(...)-127 ); 

• the maximum amplitude of 255 way big; you'll destroy whole image , noise. try few different amplitudes, a, so:

  noisyimage = max(0, min(255, originalimage + randi(a, ...)-round(a/2) ); 

• the uniform distribution randi uses not source of noise; you'd want other distribution. use normal distribution:

  uint8(a*randn(...)-round(a/2)) 

  or gamma:

  uint8(a*randg(...)-round(a/2)) 

  etc.

now, should started :)