Remove periodic noise pattern from image

Hi,

I have a set of images taken from a video which all have a regular noise pattern which I wish to remove.

You can see this pattern clearly in pattern.jpg which I attached. This is a small segment cropped from the background of image_pattern.jpg. I used imadjust on it to make the pattern visible. If you look closely at image_pattern you can see the pattern I wish to remove without losing details/sharpness in the image.

I've also attached the log of the magnitude of the fft as it seems this may be helpful in solving this problem. I obtained this with the following code:

% Compute the 2D fft. (I is image_pattern)

frequencyImage = fftshift(fft2(I));

% Take log magnitude so we can see it better in the display.

amplitudeImage = log(abs(frequencyImage));

minValue = min(min(amplitudeImage))

maxValue = max(max(amplitudeImage))

figure(3), imshow(amplitudeImage, []);

Please let me know if there are any easy solutions to this problem, thanks.

Answer 1

[Image Analyst]

I have a demo for that, attached.

[KAAN AYKUT KABAKÇI]

What is the name of this filter? Low pass, band pass or high pass?

[Image Analyst]

That would be a band pass. For a low pass, the central portion would be erased. For a high pass the outer portion of the square would be erased. Since I only erased some internal areas, it's passing stuff in the "band" outside those erase areas.

[Ali Gouhar]

How i can remove remove these diagonal lines from the image?

[Image Analyst]

Ali, did you try to apply my demo to your image? If so, you forgot to attach your code.

If it's not exact enough, I'd suggest you try a modified median filter where you create a mask where the gray lines gray values are, then replace those with the median

mask = grayImage == grayLineGrayLevel
subplot(2, 2, 1);
imshow(mask);
filteredImage = medfilt2(grayImage, [7, 7]);
subplot(2, 2, 2);
imshow(filteredImage);
grayImage(mask) = filteredImage(mask);
subplot(2, 2, 3);
imshow(grayImage);

If it's still not right, start your own question.

[Dominic Rockas]

is the fft_filter.m the same as a notch filter?