morp.m

形态滤波

function main(void)
I=imread('99','bmp');
J=im2double(I);
f=J(:,:,1);
imshow(f,[]);
b=zeros(3,3);
%b=[-2,-2,-2;-2,-2,-2;-2,-2,-2];

%x  = 0.1*randn(1,500); % Input to the filter
x = f([100:132],[100:132]);
figure,imshow(x,[]);
b  = fir1(31,0.5);     % FIR system to be identified
d  = filter(b,1,x);    % Desired signal
w0 = zeros(1,32);      % Intial filter coefficients
k0 = 0.5*eye(32);      % Initial state error correlation matrix
qm = 2;                % Measurement noise covariance
qp = 0.1*eye(32);      % Process noise covariance  
s = initkalman(w0,k0,qm,qp);
[y,e,s] = adaptkalman(x,d,s);
figure,imshow(y,[]);
stem([b.',s.coeffs.']);
legend('Actual','Estimated');
title('System Identification of an FIR filter via Kalman Filter');
grid on;

[rs,cs]=size(f);
%g=zeros(rs,cs);
g=erode(f,b);
%figure,imshow(g,[]);
g=dilate(g,b);
figure,imshow(g,[]);
g=erode(f,b);
g=f-g;
figure,imshow(g,[]);
figure,mesh(g);
%******************************************************************
function g=erode(f,b)
m=size(b,1);
m=fix(m/2);
[rs,cs]=size(f);
g=zeros(rs,cs);
for i=1:rs;
    for j=1:cs;
      tmin=1;
      for s=-m:m;
            for t=-m:m;
                if(i+s>0&&i+s<rs+1&&j+t>0&&j+t<cs+1) 
                    temp=max(0,f(i+s,j+t)-b(s+m+1,t+m+1));
                    tmin=min(tmin,temp);
                end
            end
      end
      g(i,j)=tmin;
    end
end


%******************************************************************

function g=dilate(f,b)
m=size(b,1);
m=fix(m/2);
[rs,cs]=size(f);
g=zeros(rs,cs);
for i=1:rs;
    for j=1:cs;
      tmax=0;
      for s=m:-1:-m;
            for t=m:-1:-m;
                if(i-s>0&&i-s<rs+1&&j-t>0&&j-t<cs+1) 
                    temp=min(1,f(i-s,j-t)+b(s+m+1,t+m+1));
                    tmax=max(tmax,temp);
                end
            end
      end
      g(i,j)=tmax;
    end
end     
%******************************************************************