matlab图像分割（以草莓为例），区域生长，区域分割

分割结果有点粗糙，凑合着看，原始代码是来自冈萨雷斯的数字图像处理，需要电子书可以后台询问。

区域生长

regiongrow函数，要新建一个脚本：

function [g, NR, SI, TI] = regiongrow(f, S, T)

%REGIONGROW Perform segmentation by region growing.

% [G, NR, SI, TI] = REGIONGROW(F, SR, T). S can be an array (the

% same size as F) with a 1 at the coordinates of every seed point

% and 0s elsewhere. S can also be a single seed value. Similarly,

% T can be an array (the same size as F) containing a threshold

% value for each pixel in F. T can also be a scalar, in which

% case it becomes a global threshold.

%

% On the output, G is the result of region growing, with each

% region labeled by a different integer, NR is the number of

% regions, SI is the final seed image used by the algorithm, and TI

% is the image consisting of the pixels in F that satisfied the

% threshold test.

% Copyright 2002-2004 R. C. Gonzalez, R. E. Woods, & S. L. Eddins

% Digital Image Processing Using MATLAB, Prentice-Hall, 2004

% $Revision: 1.4 $ $Date: 2003/10/26 22:35:37 $

f = double(f);

% If S is a scalar, obtain the seed image.
f=im2gray(f);

if numel(S) == 1

SI = f == S;

S1 = S;

else

% S is an array. Eliminate duplicate, connected seed locations

% to reduce the number of loop executions in the following

% sections of code.

SI = bwmorph(S, 'shrink', Inf);

J = find(SI);

S1 = f(J); % Array of seed values.

end

TI = false(size(f));

for K = 1:length(S1)

seedvalue = S1(K);

S = abs(f - seedvalue) <= T;

TI = TI | S;

end

% Use function imreconstruct with SI as the marker image to

% obtain the regions corresponding to each seed in S. Function

% bwlabel assigns a different integer to each connected region.

[g, NR] = bwlabel(imreconstruct(SI, TI));

主函数，丢另一个脚本里：

clc
clear
f = imread('E:\数字图像\第二次作业\草莓.jpg');
subplot(221),imshow(f);
title('原始图像');
%函数regiongrow返回的NR为是不同区域的数目,参数SI是一副含有种子点的图像
%TI是包含在经过连通前通过阈值测试的像素
f=im2gray(f);
[g,NR,SI,TI]=regiongrow(f,100,55);%种子的像素值为100,55为阈值
subplot(222),imshow(SI);
title('种子点的图像');
subplot(223),imshow(TI);
title('所有通过阈值测试的像素');
subplot(224),imshow(g);
title('对种子点进行8连通分析后的结果');

区域分割：

splitmerge函数，新建脚本：

function g=splitmerge(f,mindim,fun)%f是待分割的原图，mindim是定义分解中所允许的最小的块，必须是2的正整数次幂
f= rgb2gray(f); % 如果 f 是 RGB 彩色图像
 Q=2^nextpow2(max(size(f)));
 [M,N]=size(f);
 padsize = max([Q-M, Q-N], 0);
 f=padarray(f,padsize,'post');%：填充图像或填充数组。f是输入图像，输出是填充后的图像，先将图像填充到2的幂次以使后面的分解可行
 %然后是填充的行数和列数，post：表示在每一维的最后一个元素后填充,B = padarray(A,padsize,padval,direction)
 %不含padval就用0填充,Q代表填充后图像的大小。
 S=qtdecomp(f,@split_test,mindim,fun);%S传给split_test，qtdecomp divides a square image into four
% equal-sized square blocks, and then tests each block to see if it
% meets some criterion of homogeneity. If a block meets the criterion,
% it is not divided any further. If it does not meet the criterion,
% it is subdivided again into four blocks, and the test criterion is
% applied to those blocks. This process is repeated iteratively until
% each block meets the criterion. The result can have blocks of several
% different sizes.S是包含四叉树结构的稀疏矩阵，存储的值是块的大小及坐标，以稀疏矩阵形式存储
 Lmax=full(max(S(:)));%将以稀疏矩阵存储形式存储的矩阵变换成以普通矩阵（full matrix）形式存储，full，sparse只是存储形式的不同
 g=zeros(size(f));
 MARKER=zeros(size(f));
 
 for k=1:Lmax
     [vals,r,c]=qtgetblk(f,S,k);%vals是一个数组，包含f的四叉树分解中大小为k*k的块的值，是一个k*k*个数的矩阵，
     %个数是指S中有多少个这样大小的块，f是被四叉树分的原图像，r，c是对应的左上角块的坐标如2*2块，代表的是左上角开始块的坐标
     
     if ~isempty(vals)
         for I=1:length(r)
             xlow=r(I);
             ylow=c(I);
             xhigh=xlow+k-1;
             yhigh=ylow+k-1;
             region=f(xlow:xhigh,ylow:yhigh);%找到对应的区域
             flag=feval(fun,region);%evaluates the function handle, fhandle,using arguments x1 through xn.执行函数fun，region是参数
             if flag%如果返回的是1，则进行标记
                 g(xlow:xhigh,ylow:yhigh)=1;%然后将对应的区域置1
                 MARKER(xlow,ylow)=1;%MARKER矩阵对应的左上角坐标置1
             end
         end
     end
 end
 
 g=bwlabel(imreconstruct(MARKER,g));%imreconstruct默认2D图像8连通，这个函数就是起合的作用
 g=g(1:M,1:N);%返回原图像的大小
end
%%%%%%%%%%%%%%%%%%%%%%%%%%%%

function v=split_test(B,mindim,fun)
 K=size(B,3);%B就是qtdecomp函数传过来的，代表当前size(B,3)返回的是B的层数，就是B是几维的，这里实际上就是有几个B这样大小的图像块
 %这句代码的意思是从qtdecomp函数传过来的B，是当前分解成的K块的m*m的图像块，K表示有多少个这样大小的图像块
 v(1:K)=false;
 for I=1:K
     quadregion=B(:,:,I);
     if size(quadregion,1)<=mindim%如果分的块的大小小于mindim就直接结束
         v(I)=false;
         continue
     end
     flag=feval(fun,quadregion);%quadregion是fun函数的参数
     if flag%如果flag是1，代表需要再分
         v(I)=true;%这里就相当于split_test是起一个调用predicate的作用，返回的就是ppredicate的值
     end
 end
end
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

predicate函数，需要新建脚本：

function flag = predicate( region )
sd = std2(region);
m = mean2(region);
flag = (sd > 10) & (m > 0) & (m < 255);
end

主函数，需要新建脚本：

clc
clear
f = imread('E:\数字图像\第二次作业\草莓.jpg');
subplot(231),imshow(f);
title('区域分割原始图像');
g32=splitmerge(f,32,@predicate);%32代表分割中允许最小的块
subplot(232),imshow(g32);
title('mindim为32时的分割图像');
g16=splitmerge(f,16,@predicate);%32代表分割中允许最小的块
subplot(233),imshow(g16);
title('mindim为16时的分割图像');
g8=splitmerge(f,8,@predicate);%8代表分割中允许最小的块
subplot(234),imshow(g8);
title('mindim为8时的分割图像');
g4=splitmerge(f,4,@predicate);%4代表分割中允许最小的块
subplot(235),imshow(g4);
title('mindim为4时的分割图像');
g2=splitmerge(f,2,@predicate);%2代表分割中允许最小的块
subplot(236),imshow(g2);
title('mindim为2时的分割图像');

然后是分水岭分割，这个效果相比前两个来说差一点，但是代码简单：

I=imread('E:\数字图像\第二次作业\草莓.jpg');
g=im2bw(I,graythresh(I));
figure(1),subplot(321),imshow(g);
gc=~g;
subplot(322),imshow(gc);
d=bwdist(gc);
figure(1),subplot(323),imshow(d);
L=watershed(-d);
figure(1),subplot(324),imshow(L);
w=L==0;
g2=g&~w;
figure(1),subplot(325),imshow(g2);
title('分割图');
%距离变换分水岭分割

结果大概是这样：

                                     原始图

区域生长：

区域分割：

分水岭：