sector_norm.m

本程序是在matlab的开发环境下

function [disk,vector] = sector_norm( image , mode )
% Modified by Luigi Rosa
global immagine n_bands h_bands n_arcs h_radius h_lato n_sectors matrice

% N=175 size of cropped image (175 x 175)
N=h_lato;
% Number of sectors
M=n_sectors+2;

size_m=N*N;


mean_s=zeros(M,1);
varn_s=zeros(M,1);
num_s=zeros(M,1);
image1=zeros(h_lato);
Mo=100;
Vo=100;
%-------------- ------------------------------- -----------------  -----------
%----------------------------    MODE 0  -------------------------------------
if mode==0
    for ( i=1:1:size_m)
        %tmp=whichsector(i);
        tmp=matrice(i);
        tmp=tmp+1;
        
        mean_s(tmp)= mean_s(tmp)+image(i);
        num_s(tmp)=num_s(tmp)+1;
        
    end
    for (i=1:1:M)
        mean_s(i)=mean_s(i)/num_s(i);
    end
    for ( i=1:1:size_m)
        tmp=matrice(i);
        %tmp=whichsector(i);
        tmp=tmp+1;
        
        varn_s(tmp)= varn_s(tmp) + (image(i)- mean_s(tmp))^2;
        
    end
    for (i=1:1:M)
        varn_s(i)= varn_s(i) / num_s(i);
    end
    for ( i=1:1:size_m)
        tmp=matrice(i);
        %tmp=whichsector(i);
        tmp=tmp+1;
        if (abs(varn_s(tmp))>1)
            if ((image(i) - mean_s(tmp))<0)
                image1(i)=Mo - (Vo/varn_s(tmp)*((image(i) - mean_s(tmp))^2))^0.5;               
            else
                image1(i)=Mo + (Vo/varn_s(tmp)*((image(i) - mean_s(tmp))^2))^0.5;
            end
        else
            image1(i)=Mo;
        end
    end    
    disk=image1;
    vector=varn_s;   
end
%-------------- ------------------------------- -----------------  -----------
%----------------------------    MODE 1  -------------------------------------
if mode==1
    for ( i=1:1:size_m)
        %tmp=whichsector(i);
        tmp=matrice(i);
        tmp=tmp+1;
        
        mean_s(tmp)= mean_s(tmp)+image(i);
        num_s(tmp)=num_s(tmp)+1;
        
    end
    for (i=1:1:M)
        mean_s(i)=mean_s(i)/num_s(i);
    end
    for ( i=1:1:size_m)
        tmp=matrice(i);
        %tmp=whichsector(i);
        tmp=tmp+1;
        
        varn_s(tmp)= varn_s(tmp) + (image(i)- mean_s(tmp))^2;
        
    end
    for (i=1:1:M)
        varn_s(i)= varn_s(i) / num_s(i);
    end
    for (i=1:1:size_m)
        tmp=matrice(i);
        %tmp=whichsector(i);
        tmp=tmp+1;			
        image1(i)=varn_s(tmp);
    end
    vettore=zeros(M,1);
    for ( i=1:1:size_m)
        tmp=matrice(i);
        %tmp=whichsector(i);
        tmp=tmp+1;
        
        vettore(tmp)= vettore(tmp) + abs(image(i)- mean_s(tmp));
        
    end
    for (i=1:1:M)
        vettore(i)=vettore(i)/num_s(i);
    end      
    disk=image1;
    vector=vettore;    
end