1、读取图像：%Step 1: Read ImageI1 = imread('bag.png');I = rgb2gray(I1);figureimshow(I);
2、创建纹理图像使用纹理函数entropyfilt()创建纹理图像，返回与输入图像大小相同，每个像素值是输入图像相应像素值邻域的熵值；使用比例函数rescale()变成double图像；%Step 2: Create Texture ImageE = entropyfilt(I);%Eim = rescale(E);%figureimshow(Eim);
3、为底层纹理创建二值掩膜兔脒巛钒图像①分割后的图像目标区域显示为白色；②与原图像相比，图像顶部的纹理被过度分割，而底部纹理则以一个整体被分割出来；③使用bwareaopen函数提取图像的底部纹理；④分遥菇障彝割图像的边界并不光滑，并且含有很多孔洞，使用imclose函数对图像执行形态学关操作；使用imfill函数对图像中的孔洞进行填充；%Step 3: Create Rough Mask for the Bottom TextureBW1 = imbinarize(Eim,0.8);%阈值分割figureimshow(BW1);imwrite(BW1,'thresholdimg1.png');BWao = bwareaopen(BW1,2000);%开运算，连通域imshow(BWao)nhood = true(11);closeBWao = imclose(BWao,nhood);%闭合孔洞imshow(closeBWao)roughMask = imfill(closeBWao,'holes');%孔洞填充imshow(roughMask)
4、使用掩膜图像分割出上层图像%Step 4: Use Rough Mask to Segment the Top TextureI2 = I稆糨孝汶;I2(roughMask) = 0;%掩膜操作imshow(I2);imwrite(I2,'segmentimg1.png');E2 = entropyfilt(I2);%重计算纹理图像E2im = rescale(E2);imshow(E2im)BW2 = imbinarize(E2im);%自适应阈值分割imshow(BW2);mask2 = bwareaopen(BW2,1000);imshow(mask2);
5、提取底层、上层图像和绘制边界：%Step 5: Display Segmentation Resultstexture1 = I;texture1(~mask2) = 0;%上层图像imshow(texture1)texture2 = I;texture2(mask2) = 0;%底层图像figureimshow(texture2)imwrite(texture2,'bottomimg.png');boundary = bwperim(mask2);segmentResults = I;segmentResults(boundary) = 255;%绘制边界imshow(segmentResults)
6、除了使用entropyfilt函数，还可以使用标准差stdfilt()和rangefilt()来达到类似的分离效果;% Step 6:Using Other Texture Filters in SegmentationS = stdfilt(I,nhood);imshow(rescale(S));imwrite(rescale(S),'stdfiltimg.png');R = rangefilt(I,ones(5));imshow(R);imwrite(R,'rangefiltimg.png');
7、【注】：matlab rescale()函数：function y = rescale(x,a,b)% rescale - rescale data in [a,b]% y = rescale(x,a,b);% Copyright (c) 2004 Gabriel Peyr?if nargin<2 a = 0;endif nargin<3 b = 1;endm = min(x(:));M = max(x(:));y = (b-a) * (x-m)/(M-m) + a;