can i have a code to filter in the frequency domain that is capable of attenuating the noise in the image ?

The Image Below is corrupted with periodic noise , ;;

i want a filter in the frequency domain for attenuating that noise

http://tinypic.com/view.php?pic=2rh7lz7&s=5

Note : i saw alot of filters that adds ripples to an image then it removes it. so i want a filter to deal with this corrupted image without adding ripples to it , because it already has an perioidc noise

Answer 1

[Youssef Khmou]

hi lona,

there is similar question that was solved yesterday, here is the link : http://www.mathworks.com/matlabcentral/answers/82036-frequency-domain-filter-problem

However here is my suggestion :

Attenuating the noise means you need to design low pass filter, try this method :

 h=ones(3)/9;

You can see the frequency response of the filter h :

 figure, freqz2(h,32,32);

you take an image corrupted by AWGN :

 I=imnoise(im2double(imread('circuit.tif')),'Gaussian');
 X=filter2(h,I);
 figure, subplot(1,2,1), imshow(I), title(' noisy,v=0.05');
        subplot(1,2,2), imshow(X), title(' filtered, Low-pass');

[mecheal]

i saw it , but the slution wasn't the correct one , i need another solution because my image already has a periodic noise , so whay i've to add ripples to it !

Answer 2

[Image Analyst]

Come on Iona. You don't have to add ripples to it. Don't you realize, from the extensive comments in my demo, that adding the ripples was just to get some kind of starting image with periodic noise in it??? It's not part of the solution. You're starting with an image that already has periodic noise in it - evidently the same image that your classmate "Maria joe" is using. So obviously you skip the part of the demo that creates a sample image for you -- you just read in your image and go from there. And, needless to say (or maybe not so needless), your spikes will be in different locations than my demo so you'll need to change where you zero out the spikes.

[mecheal]

i'm rely srry for saying that , can u prepare it for me , because i try to do that but the result was the same of the input one ! , can u help of making a band reject filter or notch filter to my image , there is any demo for doing that???