[已答复] 用这种SNR模型，为什么算出来的SNR是1

一年四季 发表于 2016-4-22 20:41:36

1971

自编了一个5*5模板的中值滤波处理图像的程序。算出来的信噪比SNR为什么是1？
%**********************************%
%*****传统中值滤波算法的一种实现*****%
%**********************************%
clc,clear;
I0=imread('camera.jpg');
I0=rgb2gray(I0);
I=imnoise(I0,'salt & pepper',0.02);
figure(1),imshow(I0);title('原始图像');
figure(2),imshow(I);title('加了椒盐噪声的图像');
f1=I;
[m,n]=size(I);
Muban=zeros(1,25);%采用了5*5模板进行中值滤波
for i =3:m-2
for j =3:n-2
      k=1;Muban(1,k)=I(i,j);
      k=2;Muban(1,k)=I(i,j-1);
      k=3;Muban(1,k)=I(i,j-2);
      k=4;Muban(1,k)=I(i,j+1);
      k=5;Muban(1,k)=I(i,j+2);
      k=6;Muban(1,k)=I(i-1,j);
      k=7;Muban(1,k)=I(i-1,j-1);
      k=8;Muban(1,k)=I(i-1,j-2);
      k=9;Muban(1,k)=I(i-1,j+1);
      k=10;Muban(1,k)=I(i-1,j+2);
      k=11;Muban(1,k)=I(i-2,j);
      k=12;Muban(1,k)=I(i-2,j-1);
      k=13;Muban(1,k)=I(i-2,j-2);
      k=14;Muban(1,k)=I(i-2,j+1);
      k=15;Muban(1,k)=I(i-2,1+2);
      k=16;Muban(1,k)=I(i+1,j);
      k=17;Muban(1,k)=I(i+1,j-1);
      k=18;Muban(1,k)=I(i+1,j-2);
      k=19;Muban(1,k)=I(i+1,j+1);
      k=20;Muban(1,k)=I(i+1,j+2);
      k=21;Muban(1,k)=I(i+2,j);
      k=22;Muban(1,k)=I(i+2,j-1);
      k=23;Muban(1,k)=I(i+2,j-2);
      k=24;Muban(1,k)=I(i+2,j+1);
      k=25;Muban(1,k)=I(i+2,j+2);
      I(i,j)=median(Muban);
end
end
for i = 3:m-2 %处理第3到底m-2行的每一行开头和结尾的两个元素。把模板中心的像素灰度值分别赋给它们
I(i,1)=I(i,3);
I(i,2)=I(i,3);
I(i,n)=I(i,n-2);
I(i,n-1)=I(n-2);
end
I(1,:)=I(3,:);%处理头两行和末尾两行
I(2,:)=I(3,:);
I(m,:)=I(m-2,:);
I(m-1,:)=I(m-2,:);
figure(3),imshow(I);title('处理图像');
%***************计算处理后与处理前图像的信噪比**************************
sum1=0;sum2=0;clear I0;
for i=1:m
for j=1:n
      sum1=sum1+I(i,j)^2;
      sum2=sum2+(I(i,j)-f1(i,j))^2;
end
end
SNR=sum1/sum2;