AveragingAttack.m

clc;
close all;
clear all;
p=10;
chtimg=zeros(256,256);
notr=input('Enter no of traitors');
for i=1:notr
    traitor(i)=input('Enter Traitor');
end
ti=0
for m=1:p
    str = int2str(m);
    str = strcat('E:\M Tech\Semester 4\Project Work\watermarked images\',str,'.bmp');
    a = imread(str);
    b=size(a); 
    c=size(find(traitor-m));
    sizc=c(2);
    if (notr~=sizc)
        chtimg=chtimg+double(a/notr);
        ti=ti+1
    end    
    b=size(a);
    for i=1:(b(1)*b(2))
        db(i,m)=a(i);
    end
end
  A=db';
%  %e=0.1;
% % beta=0;
  d=size(A,2);
  meanvector=mean(db,2);
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%Calculating mean centered images
for l=1:p;
    meancenter(1:b(1)*b(2),l)=double(db(1:b(1)*b(2),l))-meanvector(1:b(1)*b(2));
end
mc=meancenter';
 test=chtimg;
 for i=1:b(1)*b(2)
     testvector(i,1)=test(i);
 end
% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% %%Finding the mean centered test  image
  meancenttest=double(testvector)-meanvector;
  mct=meancenttest';
% % %k=achlioptasreduceddimension(n,e,beta);
for t=1:10
k=1500;
  R=achlioptasRandomMatrix(d,k);
  r=(1/sqrt(k))*R;
%load AveragingAttack r
projectedImages = [];
for i = 1 : p
    temp = mc(i,:)*r; % projection of centered images into facespace
    projectedImages = [projectedImages; temp]; 
end
% % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% % %%projecting the test image as span of eigen faces
  projectedTestImage = mct*r; % Test image feature vector
% % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% % %%Finding the eucledian distance between test image and data base images
  for i=1:p
      euc(i,t)=(norm(projectedImages(i,:)-projectedTestImage))^2;
  end
end
eu=mean(euc,2);
(eu./max(eu))'
% % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% % %% Image with the minimum eucledian distance is the recognised image
  %[mindis arg_min]=min(euc);
% % %%%%%%%%%%%%%%%%%%%%%%%%%%%
%   arg_min
%   mindis

  %no=euc./max(euc)