prototype pattern and adjoint pattern.m

协同神经网络的程序

R1=imread('xsi','bmp');
Y(1)=mean(mean(R1));
B1=(R1(:))';
v(1,:)=double(B1);
v(1,:)=v(1,:)-mean(v(1,:));
v(1,:)=v(1,:)/norm(v(1,:));

R2=imread('xtuo','bmp');
Y(2)=mean(mean(R2));
B2=(R2(:))';
v(2,:)=double(B2);
v(2,:)=v(2,:)-mean(v(2,:));
v(2,:)=v(2,:)/norm(v(2,:));

R3=imread('xsan','bmp');
Y(3)=mean(mean(R3));
B3=(R3(:))';
v(3,:)=double(B3);
v(3,:)=v(3,:)-mean(v(3,:));
v(3,:)=v(3,:)/norm(v(3,:));


%we get the value of pinv v
VV = (pinv(v))';

v1=VV(1,:);
v1 = reshape(v1,15,15);
v1 = mat2gray(v1*255);
imshow(v1);

v2=VV(2,:);
v2 = reshape(v2,15,15);
v2 = mat2gray(v2*255);
imshow (v2);

v3=VV(3,:);
v3 = reshape(v3,15,15);
v3 = mat2gray(v3*255);
imshow (v3);

for i=1:3
   for j=1:3 
      a = v(i,:)*VV(j,:)'
   end    
end

%Q is the row parament
R4=imread('shapez','bmp');
Y(4)=mean(mean(R4));
Q=double(R4(:));
Q=Q-mean(Q);
Q=Q/norm(Q);

%A is the order parameter
for k=1:3
   A(k)=VV(k,:)*Q;
end