Noise Removing of an image matlab Code

Hi! Everyone i am new in matlab . I tried many process in matlab but can not get my desire output. Can anyone help me with some code that can remove image noise. The image is bellow. Please give me code to remove noise.. I need only white part and the other part will be black. Any help will be appreciated....... Thanks in advance...

[Image Analyst]

Do you mean the white pixels as shown above, or the white part of the eye? And post your original gray scale or color image, and code to arrive at the binary image you showed, not a screenshot of a figure window. A screenshot is not that useful as far as trying anything goes.

[Asif Hasan]

Sorry sir, Its my first time so. And i got this figure using the adaptive clustering process the code is

function [lb,center] = adaptcluster_kmeans(im)

% This code is written to implement kmeans clustering for segmenting any % Gray or Color image. There is no requirement to mention the number of cluster for % clustering. % IM - is input image to be clustered. % LB - is labeled image (Clustered Image). % CENTER - is array of cluster centers. % Execution of this code is very fast. % It generates consistent output for same image.

if size(im,3)>1 [lb,center] = ColorClustering(im); % Check Image is Gray or not. else [lb,center] = GrayClustering(im); end

function [lb,center] = GrayClustering(gray) gray = double(gray); array = gray(:); % Copy value into an array. % distth = 25; i = 0;j=0; % Intialize iteration Counters. tic while(true) seed = mean(array); % Initialize seed Point. i = i+1; %Increment Counter for each iteration. while(true) j = j+1; % Initialize Counter for each iteration. dist = (sqrt((array-seed).^2)); % Find distance between Seed and Gray Value. distth = (sqrt(sum((array-seed).^2)/numel(array)));% Find bandwidth for Cluster Center. % distth = max(dist(:))/5; qualified = dist<distth;% Check values are in selected Bandwidth or not. newseed = mean(array(qualified));% Update mean.

        if isnan(newseed) % Check mean is not a NaN value.
            break;
        end

        if seed == newseed || j>10 % Condition for convergence and maximum iteration.
            j=0;
            array(qualified) = [];% Remove values which have assigned to a cluster.
            center(i) = newseed; % Store center of cluster.
            break;
        end
        seed = newseed;% Update seed.
    end

    if isempty(array) || i>10 % Check maximum number of clusters.
        i = 0; % Reset Counter.
        break;
    end

end toc

center = sort(center); % Sort Centers. newcenter = diff(center);% Find out Difference between two consecutive Centers. intercluster = (max(gray(:)/10));% Findout Minimum distance between two cluster Centers. center(newcenter<=intercluster)=[];% Discard Cluster centers less than distance.

% Make a clustered image using these centers.

vector = repmat(gray(:),[1,numel(center)]); % Replicate vector for parallel operation. centers = repmat(center,[numel(gray),1]);

distance = ((vector-centers).^2);% Find distance between center and pixel value. [~,lb] = min(distance,[],2);% Choose cluster index of minimum distance. lb = reshape(lb,size(gray));% Reshape the labelled index vector.

function [lb,center] = ColorClustering(im)

im = double(im); red = im(:,:,1); green = im(:,:,2); blue = im(:,:,3);

array = [red(:),green(:),blue(:)]; % distth = 25; i = 0;j=0; tic while(true)

    seed(1) = mean(array(:,1));
    seed(2) = mean(array(:,2));
    seed(3) = mean(array(:,3));

    i = i+1;
    while(true)
        j = j+1;

        seedvec = repmat(seed,[size(array,1),1]);

        dist = sum((sqrt((array-seedvec).^2)),2);

         distth = 0.25*max(dist);
        qualified = dist<distth;

        newred = array(:,1);
        newgreen = array(:,2);
        newblue = array(:,3);

        newseed(1) = mean(newred(qualified));
        newseed(2) = mean(newgreen(qualified));
        newseed(3) = mean(newblue(qualified));

        if isnan(newseed)
            break;
        end

        if (seed == newseed) | j>10
            j=0;
            array(qualified,:) = [];
            center(i,:) = newseed;
            %             center(2,i) = nnz(qualified);
            break;
        end
        seed = newseed;
    end

    if isempty(array) || i>10
        i = 0;
        break;
    end

end toc centers = sqrt(sum((center.^2),2)); [centers,idx]= sort(centers);

while(true) newcenter = diff(centers); intercluster =25; %(max(gray(:)/10)); a = (newcenter<=intercluster); % center(a,:)=[]; % centers = sqrt(sum((center.^2),2)); centers(a,:) = []; idx(a,:)=[]; % center(a,:)=0; if nnz(a)==0 break; end

end center1 = center; center =center1(idx,:); % [~,idxsort] = sort(centers) ; vecred = repmat(red(:),[1,size(center,1)]); vecgreen = repmat(green(:),[1,size(center,1)]); vecblue = repmat(blue(:),[1,size(center,1)]);

distred = (vecred - repmat(center(:,1)',[numel(red),1])).^2; distgreen = (vecgreen - repmat(center(:,2)',[numel(red),1])).^2; distblue = (vecblue - repmat(center(:,3)',[numel(red),1])).^2;

distance = sqrt(distred+distgreen+distblue); [~,label_vector] = min(distance,[],2); lb = reshape(label_vector,size(red)); %

and the color image is

I need to segment the sclera part from the image. I am doing sclera recognition project. If you have any matlab code for this please help. And thanks for reply.

[Image Analyst]

Please just attach the m-file with the paper clip icon. Then read this http://www.mathworks.com/matlabcentral/answers/13205-tutorial-how-to-format-your-question-with-markup

Answer 1

[Image Analyst]

Look here for algorithms : http://www.visionbib.com/bibliography/contents.html

Search for sclera: http://www.visionbib.com/bibliography_search.html

[Image Analyst]

See attached demo (test.m at the very bottom), taken from my file exchange, adapted to find white. it will produce this image (on its second screen):

You can see from this 3D color gamut:

that kmeans would not necessarily be great for color segmentation, and especially not after you've converted to gray scale. So I would not use the approach you mentioned.

[Asif Hasan]

Thank you so much for your help sir.

[Asif Hasan]

Sir,

Would you please brief me about the code that you gave me so that i can understand.When i run following image through the code i got a white section upside the eye but i can't understand how to modify it. So give me some idea so that i do my work and sir how do you measure the red color?

[Image Analyst]

Yes, some of the eyelid is white, and so it finds that region. Try adjusting some of the parameters or look for successful published algorithms here: http://www.visionbib.com/bibliography/contents.html

[Asif Hasan]

@Image Analist Sir, You provided me a code that can segment sclera from eye image

but when i input this image it can't give me the result. What should i do?

[Image Analyst]

Keep working on it. Did you see the link to papers from people have have presumably solved it? It looks like color alone won't solve it if you need to find sclera that is reddish so you may have to combine with other methods like spatial location or edge detection. Maybe you have to locate the iris first and look outwards from there. I can see that it could be a very challenging problem to automatically get every possible pathological image absolutely correct. As a last resort you may have to resort to using imfreehand(). Let me know if you want a demo of that. The good thing is that if you resort to doing it manually, there aren't many images since each patient will have only a few images at most, not hundreds or thousands.

[Asif Hasan]

Thanks for the reply . But which one should i follow first for this kind of problem , there are many links. @Image Analist

[Image Analyst]

Follow the ones that sound most promising. I have not read all the image processing papers in that narrow field, or even any of them, so I don't know which are any good. You'll have to figure that out on your own.