Matlab数字图像处理基础【9】

来源：互联网发布：云服务器限制端口编辑：程序博客网时间：2024/04/30 22:49

第六章图像分割

实例：灰度图像的阈值分割

clc;clear;%读入图像A = imread('shape.jpg');figure(1),subplot(2,2,1),imshow(A),title('原图像');%获取灰度直方图，根据直方图人工选取阈值h = imhist(A);%h = imhist(A) / numel(A);h1 = h(1:10:256);horz = 1:10:256;figure(2),bar(horz, h1),axis([0 255 0 15000]),title('直方图');%人工设定阈值进行图像分割thr1 = 100;[m, n] = size(A);for i = 1:m    for j = 1:n        if A(i,j) < thr1            A1(i,j) = 0;        else            A1(i,j) = 1;        end    endendfigure(1),subplot(2,2,2),imshow(A1),title('人工设定阈值');%使用迭代法选取阈值进行图像分割%计算阈值初始值thr2 = 0.5 * (double(min(A(:))) + double(max(A(:))));%设定循环终止条件done = false;%循环计算阈值和分割图像while ~done    %获取二值分割结果    A2 = A >= thr2;    %计算新阈值    tnext = 0.5 * (mean(A(A2)) + mean(A(~A2)));    %判断循环是否终止    done = abs(thr2 - tnext) < 0.5;    %阈值更新    thr2 = tnext;endfigure(1),subplot(2,2,3),imshow(A2),title('迭代选取阈值');%使用Otsu方法进行阈值选取和图像分割%使用Otsu方法获取阈值thr3 = graythresh(A) * 255;%使用获得的阈值进行阈值分割[m, n] = size(A);for i = 1:m    for j = 1:n        if A(i,j) < thr3            A3(i,j) = 0;        else            A3(i,j) = 1;        end    endendfigure(1),subplot(2,2,4),imshow(A3),title('Otsu阈值选取');

实验结果：

0 0