car2.m

%% car2(): Feautures extraction function
% This function creates a set of ideal filters and applies them to the
% input images. Aafter that, the energy is calculated for each resulting
% filtered image 
% .The output is a features vector with the same size as the rings number
function features=car2(im)

%%
% Randomly select a pixel within the image range and create a new subimage
% _cort_
ima=imread(im);
[l1 l2 l3]=size(ima);
l=randi(l1-100);
r=randi(l2-100);
cort=ima((l:l+99),(r:r+99));


[b c]=spect1(cort); % Calls the _spect1()_ function
[x,y]=size(b);

n=50; % Rings number
if x>=y
h=x/n;

end

%%
% Calculate energy in horizontal spectrum rings

Eh=[];
for i=1:n/2
    f=0;
    g=lpfilter('ideal',x,y,h*i);
    g2=1-lpfilter('ideal',x,y,h*(i-1));
    fi=b.*g.*g2;
    d=uint8(ifft2(fi));
    e=sum(d(:));
    Eh(i)=e;
%   if i<=16
% figure(4),subplot(4,4,i),imshow(d),title(e);
%     end

end

%%
% Calculate energy in vertical spectrum rings adn display the reconstructed
% image after the applied filter

Ev=[];
for i=1:n/2

    g=lpfilter('ideal',x,y,h*i);
    g2=1-lpfilter('ideal',x,y,h*(i-1));
    fi=c.*g.*g2;

    d=uint8(ifft2(fi));
    e=sum(d(:));
%   if i<=16
% figure(4),subplot(4,4,i),imshow(d),title(e);
%     end
   
    Ev(i)=e;
end

Eh=Eh/max(Eh);
Ev=Ev/max(Ev);

%%
% Features vector and output of the function
features(1:25)=Eh;
features(26:50)=Ev;
% figure,bar(features)