clustering image, display output

Question

Tomas 2014-3-14

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/121499-clustering-image-display-output

编辑： Tomas 2014-3-14

I have function

clc,clear all
[I,map]=imread('test','bmp');
I = ~I;
imshow(I,map);
[m n]=size(I);
P = [];
for i=1:m    
    for j=1:n        
        if I(i,j)==1
            P = [P ; i j];        
        end
    end
end
size(P);
MON=P;
      k = 3; % zadanie k
      TAZ=MON(1:k,:); % 3 centroids
      %%2 krok - pomocne premenne
  [m,n]=size(MON); %rozmery MON
  itr=1; %aktualna iteracia
  E = inf;% kvalita poslednej iter
  Enew = 0;
  PEO=0;
  PEO1=0;
    while any(E~=Enew) && PEO>=PEO1 
    Enew=E;
    PEO1=PEO;
%%4 krok - priradenie bodov do zhlukov podla tazisk 
     idx=1;
     [r,s]=size(TAZ); %rozmery TAZ
      ED=[];%euklidovska vzdialenost medzi premennou B a typickym bodom
      Z=cell(1,k);  %vytvorenie zhlukov
      z = 1;
      for i=1:m % pre pocet objektov (pre kazdy riadok MON)
          B = MON(i,:); % vloz i-ty riadok z MON do premennej B
          for j=1:r % pre kazdy riadok TAZ
              ED(j)= euklid6(TAZ(j,:),B); %euklidovska vzdialenost
              [minD, minI(idx)] = min([ED]); %minimalna hodnota a index
          end
          [e,f]=size(ED);
          for h=1:f
              if (ED(h) == ED(minI(idx)))          
              Z{h}{z}=B;
              z=z+1;
            end
          end   
          idx=idx+1;
      end
      dZ=length(Z); % dlzka Z
     %odstranenie prazdnych miest
        for i=1:size(Z,2)
          for j=length(Z{i})
            if j~=0
               Z{i}(cellfun(@isempty,Z{i})) = [];
            else continue;
            end
        end
     end
      Z(cellfun('isempty',Z))=[];%odstranenie prazdnych miest
      k=length(Z);
      minI';
%new centroid 
 TAZ_new=[];      
    for j=1:k
        TAZ_new(j,:)=tazisko6(Z(j));%vypocet noveho taziska  
    end
TAZ=TAZ_new;
E=sum(E);
TAZ=TAZ_new;    
k=length(Z);
itr = itr+1;
PEO=E; 
disp(Z);
    end
    disp('Koniec')

%% k means clustering

function [d] = tazisko6(A)
%%[d] = tazisko6(A)
%
% Tazisko jednotlivych zhlukov(centroid)
d1=length(A);
d2 =length(A{1});
     s=0; 
     for i=1:d1
         for j=1:d2
             s=s+(A{i}{j});
         end
         if d2~=1
         d=s/d2;
         else
         d=s;
         end
     end
function [ d] = euklid6( A,B )
%%[d] = euklid6(A,B)
%
% Euklidovska vzdialenost bodov
[hA,tA]=size(A);
[hB,wB]=size(B);
z=1;
        if hA==1 & hB==1
              for i=1:hA
                  d(i)=sqrt(sum((A(i,:) - B(i,:)).^2));
              end
          else
              if hA==1 & hB==1 & tA==1 & wB==1
              for i=1:hA
                  for j=1:hB
                         d(z)=sqrt(sum((A(i,:) - B(j,:)).^2));
                         z=z+1;
                      end
              end
                      else
                for i=1:hA
                  for j=1:hB
                         d(z)=sqrt(sum((A(i,:) - B(j,:)).^2));
                         z=z+1;
                      end
                end
          end
      end
end

My output is Z={1x1 cell} {1x5 cell} {1x4 cell}.

as I show the individual cell in image?

Z{1}= first cluster Z{2}= second cluster Z{3}= third cluster

each cluster must be different colors.

Thanks.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

clustering image, display output

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

clustering image, display output

0 个评论 显示 -2更早的评论隐藏 -2更早的评论

回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

0 个评论
显示 -2更早的评论隐藏 -2更早的评论