基于小波变换的模糊图像融合算法matlab仿真

function x3=func_fusion4(x1,x2); %低频小波系数绝对值较大融合策略，高频基于局部显著性准则 x1=rgb2gray(x1); x1=double(x1)/255; x2=rgb2gray(x2); x2=double(x2)/255; subplot(221),imshow('1.bmp'),title('待融合图像1'); subplot(222),imshow('2.bmp'),title('待融合图像2'); [ca1,ch1,cv1,cd1]=dwt2(x1,'db1'); [ca2,ch2,cv2,cd2]=dwt2(x2,'db1'); [row,col]=size(ca1); fun1 = inline('max(abs(x(:)))') sh1=nlfilter(ch1,[3 3],fun1); sh2=nlfilter(ch2,[3 3],fun1); sv1=nlfilter(cv2,[3 3],fun1); sv2=nlfilter(cv2,[3 3],fun1); sd1=nlfilter(cd1,[3 3],fun1); sd2=nlfilter(cd2,[3 3],fun1); for i=1:row for j=1:col %-----低频融合策略 -----% %低频取平均 % ca(i,j)=(ca1(i,j)+ca2(i,j))/2; %低频取绝对值大 if abs(ca1(i,j))>abs(ca2(i,j)) ca(i,j)=ca1(i,j); else ca(i,j)=ca2(i,j); end %-----高频融合策略 -----% if sh1(i,j)>=sh2(i,j) mhx(i,j)=1; mhy(i,j)=0; else mhx(i,j)=0; mhy(i,j)=1; end if sv1(i,j)>=sv2(i,j) mvx(i,j)=1; mvy(i,j)=0; else mvx(i,j)=0; mvy(i,j)=1; end if sd1(i,j)>=sd2(i,j) md(i,j)=1; md(i,j)=0; else md(i,j)=0; md(i,j)=1; end end end fun2 = inline('sum(x(:))') mxx=nlfilter(mhx,[3 3],fun2); mvv=nlfilter(mvx,[3 3],fun2); mdd=nlfilter(mhx,[3 3],fun2); for i=1:row for j=1:col if mxx(i,j)>=5 ch(i,j)=ch1(i,j); else ch(i,j)=ch2(i,j); end if mvv(i,j)>=5 cv(i,j)=cv1(i,j); else cv(i,j)=cv2(i,j); end if mdd(i,j)>=5 cd(i,j)=cd1(i,j); else cd(i,j)=cd2(i,j); end end end x3=idwt2(ca,ch,cv,cd,'db1');