1、corr2
A = ;
B = ;
cor = corr2(A,B)
2、通过直方图相似性比较
(参考:https://blog.csdn.net/xiaojianzhao/article/details/103785809)
如果两张图像的直方图一样,或者有极高的相似度,那么在一定程度上,可以认为这两幅图是一样的,这就是直方图比较的应用之一。通过默认的欧氏距离来比较直方图的相似性大小,值越小,两个图像的直方图越相似。
i1 = a;
i2 = b;
[c1,n]=imhist(i1);
c1=c1/size(i1,1)/size(i1,2);
[c2,n2]=imhist(i2);
c2=c2/size(i2,1)/size(i2,2);
d=pdist2(c1',c2')
3、直方图 - 巴氏距离、欧式距离、直方图相交距离
(参考:https://blog.csdn.net/shanshangyouzhiyangM/article/details/52217644)这篇博文的代码运行,会报Count1未声明,我得代码将Count1改为CountI,count2改为CountJ,不知道对不对,搞清楚了回来补笔记
I=imread('6.tif');
J=imread('06.tif');
[CountI]=imhist(I);
[CountJ]=imhist(J);
%% 巴氏距离 =0时最匹配 =1时完全不匹配%
SumI=sum(CountI);
SumJ=sum(CountJ);
%Sumup = sqrt(Count1.*Count2);
%SumDown = sqrt(Sum1*Sum2);
Sumup = sqrt(CountI.*CountJ);
SumDown = sqrt(SumI*SumJ);
Sumup = sum(Sumup);
HistDist=sqrt(1-Sumup/SumDown)
%% 欧式距离 =0时最匹配%
HistDist=sqrt(sum((CountI-CountJ).*(CountI-CountJ)))
%% 直方图相交距离 =1时最匹配 越小则越不匹配%
HistDist=sum(min(CountI,CountJ))/sum(CountI)
%% 图片和直方图显示%
figure(1);
subplot(2,2,1);imshow(I);
subplot(2,2,2);imshow(J);
subplot(2,2,3);imhist(I);
subplot(2,2,4);imhist(J);
4、结构相似性度量
(参考:https://blog.csdn.net/u010977034/article/details/82733137)
img1 =
img2 =
[ssimval, ssimmap] = ssim(img1,img2);