pic2vec.m

matlab平台上的图像处理源代码

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Main function for the fingerprint identification
% gets the character vector from the fingerprint picture
% written by zhang_rui at Chinese Academy of Science
% E-mail:[e-mail]
% 2006/6/3
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%--------------------------------------------------------------------------
% a short description:
% input a gray fingerprint picture return the character vector of it.
% The main function calls :img=preprocessing(A);[ii,jj]=centralizing(A);
% Sam=samfun(r,R,n,N1);The following list gives a brief description:
%--------------------------------------------------------------------------
%M-file list:
%   
%   preprocessing.m:(be called by main function pic2vec)
%                   preprocess the picture including:Wiener filter;
%                   meddle value filter, change a gray picture to a 
% 				     binary one and then get the skeleton.
%   centralizing.m: (be called by main function pic2vec)
%                   input the source picture output the core point of 
%                   the fingerprint.Gabor filter is used in this function,
%                   and it call gabor2d_sub.m to generate a Gabor filter.
%                   it also call varpic.m to calculate a variance picture.
%   samfun.m:       (be called by main function pic2vec)
%                   generate the sample pattern automatically, which is 
%                   used to sample the fingerprint.
%   gabor2d_sub:    (be called by centralizing.m)
%                   get gabor filter.
%   varpic.m:       (be called by centralizing.m)
%                   calculate variance picture
%--------------------------------------------------------------------------
% Globl parameter:
%
%   the following parameter is used to define the sample pattern
%   all these parameter is extremely correlate to the application
%   and can be decided by experiment
%   r=radius of smallest circle(default 10)
%   R=radius of the biggest circls(default 160)
%   n=number of the circle(default 7)
%   N1=first circle's sample N1 times(default 4)
%--------------------------------------------------------------------------

function cha=pic2vec(A)
r=10;%r=10;
R=110;%R=160;
n=5;%n=7;
N1=4;%N1=4;
%--------------------------------------------------------------------------
%call preprocessing(),findcentral(),and samfun
img=preprocessing(A);
[ii,jj]=centralizing(A);
Sam=samfun(r,R,n,N1);

%ii=250;
%jj=200;

%[Outputprint,XofCenter,YofCenter]=centralizing(img);

%--------------------------------------------------------------------------
%sampling the fingerprint
[Mp,Np]=size(img);
[Ms,Ns]=size(Sam);
img=and(img((ii-fix(Ms/2)):(ii+fix(Ms/2)),(jj-fix(Ms/2)):(jj+fix(Ms/2))),Sam);
figure;
imshow(img,[]);
%--------------------------------------------------------------------------
%generate the character vector
veclen=N1*(1+n)*n/2;
cha=zeros(1,veclen);

width=(R-r)/n;
for i=1:n
    if(i==1)
        r=r;
    else
        r=r+width;
    end
    for j=1:N1*i
        sita=2*pi*(j-1)/(N1*i);
        temp=0;
        for k=1:width
                if((img(R-fix((r+k-1)*sin(sita)),R+fix((r+k-1)*cos(sita))))==1&temp==0)
                    temp=1;        
                    cha((i-1)*i*N1/2+j)=cha((i-1)*i*N1/2+j)+1;
                else
                    if((img(R-fix((r+k-1)*sin(sita)),R+fix((r+k-1)*cos(sita))))==0)
                        temp=0;    
                    end
                end
            end
        end
    end
end
figure;
stem(cha);