Question

如何用matlab实现用小波变换对图像的纹理特征进行抽取...

RT，求代码...谢谢...

若以下回答无法解决问题，邀请你更新回答

Answer 1

%计算Tamura纹理特征
close all
clear all
clc
t0=cputime;
I = imread('F-18.bmp');
[Nx,Ny] = size(I);
Ng=256;
G=double(I);
%计算粗糙度(coarseness)
Sbest=zeros(Nx,Ny);
E0h=zeros(Nx,Ny);
E0v=zeros(Nx,Ny);
E1h=zeros(Nx,Ny);
E1v=zeros(Nx,Ny);
E2h=zeros(Nx,Ny);
E2v=zeros(Nx,Ny);
E3h=zeros(Nx,Ny);
E3v=zeros(Nx,Ny);
E4h=zeros(Nx,Ny);
E4v=zeros(Nx,Ny);
E5h=zeros(Nx,Ny);
E5v=zeros(Nx,Ny);
flag=0;

for i=1:Nx
for j=2:Ny
E0h(i,j)=G(i,j)-G(i,j-1);
end
end
E0h=E0h/2;

for i=1:Nx-1
for j=1:Ny
E0v(i,j)=G(i,j)-G(i+1,j);
end
end
E0v=E0v/2;

%图片大小必须大于4*4才能计算E1h、E1v
if (Nx<4||Ny<4)
flag=1;
end
if(flag==0)
for i=1:Nx-1
for j=3:Ny-1
E1h(i,j)=sum(sum(G(i:i+1,j:j+1)))-sum(sum(G(i:i+1,j-2:j-1)));
end
end
for i=2:Nx-2
for j=2:Ny
E1v(i,j)=sum(sum(G(i-1:i,j-1:j)))-sum(sum(G(i+1:i+2,j-1:j)));
end
end
E1h=E1h/4;
E1v=E1v/4;
end

%图片大小必须大于8*8才能计算E2h、E2v
if (Nx<8||Ny<8)
flag=1;
end
if(flag==0)
for i=2:Nx-2
for j=5:Ny-3
E2h(i,j)=sum(sum(G(i-1:i+2,j:j+3)))-sum(sum(G(i-1:i+2,j-4:j-1)));
end
end
for i=4:Nx-4
for j=3:Ny-1
E2v(i,j)=sum(sum(G(i-3:i,j-2:j+1)))-sum(sum(G(i+1:i+4,j-2:j+1)));
end
end
E2h=E2h/16;
E2v=E2v/16;
end

%图片大小必须大于16*16才能计算E3h、E3v
if (Nx<16||Ny<16)
flag=1
end
if(flag==0)
for i=4:Nx-4
for j=9:Ny-7
E3h(i,j)=sum(sum(G(i-3:i+4,j:j+7)))-sum(sum(G(i-3:i+4,j-8:j-1)));
end
end
for i=8:Nx-8
for j=5:Ny-3
E3v(i,j)=sum(sum(G(i-7:i,j-4:j+3)))-sum(sum(G(i+1:i+8,j-4:j+3)));
end
end
E3h=E3h/64;
E3v=E3v/64;
end

%图片大小必须大于32*32才能计算E4h、E4v
if (Nx<32||Ny<32)
flag=1;
end
if(flag==0)
for i=8:Nx-8
for j=17:Ny-15
E4h(i,j)=sum(sum(G(i-7:i+8,j:j+15)))-sum(sum(G(i-7:i+8,j-16:j-1)));
end
end
for i=16:Nx-16
for j=9:Ny-7
E4v(i,j)=sum(sum(G(i-15:i,j-8:j+7)))-sum(sum(G(i+1:i+16,j-8:j+7)));
end
end
E4h=E4h/256;
E4v=E4v/256;
end

%图片大小必须大于64*64才能计算E5h、E5v
if (Nx<64||Ny<64)
flag=1;
end
if(flag==0)
for i=16:Nx-16
for j=33:Ny-31
E5h(i,j)=sum(sum(G(i-15:i+16,j:j+31)))-sum(sum(G(i-15:i+16,j-32:j-31)));
end
end
for i=32:Nx-32
for j=17:Ny-15
E5v(i,j)=sum(sum(G(i-31:i,j-16:j+15)))-sum(sum(G(i+1:i+32,j-16:j+15)));
end
end
E5h=E5h/1024;
E5v=E5v/1024;
end

for i=1:Nx
for j=1:Ny
[maxv,index]=max([E0h(i,j),E0v(i,j),E1h(i,j),E1v(i,j),E2h(i,j),E2v(i,j),E3h(i,j),E3v(i,j),E4h(i,j),E4v(i,j),E5h(i,j),E5v(i,j)]);
k=floor((index+1)/2);
Sbest(i,j)=2.^k;
end
end
Fcoarseness=sum(sum(Sbest))/(Nx*Ny);
%计算对比度
[counts,graylevels]=imhist(I);
PI=counts/(Nx*Ny);
averagevalue=sum(graylevels.*PI);
u4=sum((graylevels-repmat(averagevalue,[256,1])).^4.*PI);
standarddeviation=sum((graylevels-repmat(averagevalue,[256,1])).^2.*PI);
alpha4=u4/standarddeviation^2;
Fcontrast=sqrt(standarddeviation)/alpha4.^(1/4);
%计算方向度
PrewittH=[-1 0 1;-1 0 1;-1 0 1];
PrewittV=[1 1 1;0 0 0;-1 -1 -1];
%计算横向梯度
deltaH=zeros(Nx,Ny);
for i=2:Nx-1
for j=2:Ny-1
deltaH(i,j)=sum(sum(G(i-1:i+1,j-1:j+1).*PrewittH));
end
end
for j=2:Ny-1
deltaH(1,j)=G(1,j+1)-G(1,j);
deltaH(Nx,j)=G(Nx,j+1)-G(Nx,j);
end
for i=1:Nx
deltaH(i,1)=G(i,2)-G(i,1);
deltaH(i,Ny)=G(i,Ny)-G(i,Ny-1);
end
%计算竖向梯度
deltaV=zeros(Nx,Ny);
for i=2:Nx-1
for j=2:Ny-1
deltaV(i,j)=sum(sum(G(i-1:i+1,j-1:j+1).*PrewittV));
end
end
for j=1:Ny
deltaV(1,j)=G(2,j)-G(1,j);
deltaV(Nx,j)=G(Nx,j)-G(Nx-1,j);
end
for i=2:Nx-1
deltaV(i,1)=G(i+1,1)-G(i,1);
deltaV(i,Ny)=G(i+1,Ny)-G(i,Ny);
end
%梯度向量模
deltaG=(abs(deltaH)+abs(deltaV))/2;
%梯度向量方向
theta=zeros(Nx,Ny);
for i=1:Nx
for j=1:Ny
if (deltaH(i,j)==0)&&(deltaV(i,j)==0)
elseif deltaH(i,j)==0
theta(i,j)=pi;
else
theta(i,j)=atan(deltaV(i,j)/deltaH(i,j))+pi/2;
end
end
end
theta1=reshape(theta,1,[]);
phai=0:0.0001:pi;
HD1=hist(theta1,phai);
HD1=HD1/(Nx*Ny);
HD2=zeros(size(HD1));
%定义一个阈值THRESHOLD
THRESHOLD=0.01;
for m=1:length(HD2)
if HD1(m)>=THRESHOLD
HD2(m)=HD1(m);
end
end
[c,index]=max(HD2);
phaiP=index*0.0001;
Fdirection=0;
for m=1:length(HD2)
if HD2(m)~=0
Fdirection=Fdirection+(phai(m)-phaiP)^2*HD2(m);
end
end
disp('粗糙度：');display(Fcoarseness)
disp('对比度：');display(Fcontrast)
disp('方向度：');display(Fdirection)
deltaT=cputime-t0;
display(deltaT);