How to compute median of neighboring pixels while using 3 x 3 sliding window?

I want to replace any noisy pixel with the median of neighboring pixels in MATLAB. Kindly guide what code I can use?

I have an image which contains noisy pixels, I want to replace the noisy pixels value with the median of its 8 neighbors median value.

I will be very thankful to all.

Answer 1

[Guillaume]

Use medfilt2:

filteredimage = medfilt2(yourimage, [3 3]);

Note that the current pixel is included in the median calculation, so that's 9 pixels not 8 taken into account.

[shafaq nisar]

I tried this but error shown

x = im2double(imread('Noise-01.png')); >> imshow(x); >> filteredimage = medfilt2(x, [3 3]); Error using medfilt2 Expected input number 1, A, to be two-dimensional.

Error in medfilt2>parse_inputs (line 107) validateattributes(a, ...

Error in medfilt2 (line 48) [a, mn, padopt] = parse_inputs(varargin{:});

Please guide

[Image Analyst]

It's a color image. Either do it one color channel at a time, or convert the image to grayscale.

[Guillaume]

Clearly, as the error message tells you, medfilt2 expects a 2D matrix. Therefore, your matrix must be 3D, that is a colour image. You haven't really told us what you want to do in that case. If you want to apply the median filter to each colour channel independently then you can loop over the channels, or use medfilt3 with a size of 1 in the 3rd dimension:

medfilt3(yourcolourimage, [3 3 1])