五种实现matlab边缘检测算法：
方法一：
matlab自带的edge函数：
将图片保存为lena.jpg
I=imread(‘lena.jpg‘);%提取图像 img=rgb2gray(I); [m,n]=size(img); BW1=edge(img,‘sobel‘); %用Sobel算子进行边缘检测 BW2=edge(img,‘roberts‘);%用Roberts算子进行边缘检测 BW3=edge(img,‘prewitt‘); %用Prewitt算子进行边缘检测 BW4=edge(img,‘log‘); %用Log算子进行边缘检测 BW5=edge(img,‘canny‘); %用Canny算子进行边缘检测 h=fspecial(‘gaussian‘,5);%?高斯滤波 BW6=edge(img,‘canny‘);%高斯滤波后使用Canny算子进行边缘检测 subplot(2,3,1), imshow(BW1); title(‘sobel edge check‘); subplot(2,3,2), imshow(BW2); title(‘roberts edge check‘); subplot(2,3,3), imshow(BW3); title(‘prewitt edge check‘); subplot(2,3,4), imshow(BW4); title(‘log edge check‘); subplot(2,3,5), imshow(BW5); title(‘canny edge check‘); subplot(2,3,6), imshow(BW6); title(‘gasussian&canny edge check‘);
效果如下图所示：
方法二：Laplacian算法
clear;
sourcePic=imread(‘lena.jpg‘);%图像读入
grayPic=mat2gray(sourcePic);%实现图像的矩阵归一化操作
[m,n]=size(grayPic);
newGrayPic=grayPic;
LaplacianNum=0;%经Laplacian操作得到的每个像素的值
LaplacianThreshold=0.2;%设定阈值
for j=2:m-1 %进行边界提取
for k=2:n-1
LaplacianNum=abs(4*grayPic(j,k)-grayPic(j-1,k)-grayPic(j+1,k)-grayPic(j,k+1)-grayPic(j,k-1));
if(LaplacianNum > LaplacianThreshold)
newGrayPic(j,k)=255;
else
newGrayPic(j,k)=0;
end
end
end
figure,imshow(newGrayPic);
title(‘Laplacian算子的处理结果‘)
效果图如下：
方法三：Prewitt算法
%Prewitt 算子的实现：
clear;
sourcePic=imread(‘lena.jpg‘);
grayPic=mat2gray(sourcePic);
[m,n]=size(grayPic);
newGrayPic=grayPic;
PrewittNum=0;
PrewittThreshold=0.5;%设定阈值
for j=2:m-1 %进行边界提取
for k=2:n-1
PrewittNum=abs(grayPic(j-1,k+1)-grayPic(j+1,k+1)+grayPic(j-1,k)-grayPic(j+1,k)+grayPic(j-1,k-1)-grayPic(j+1,k-1))+abs(grayPic(j-1,k+1)+grayPic(j,k+1)+grayPic(j+1,k+1)-grayPic(j-1,k-1)-grayPic(j,k-1)-grayPic(j+1,k-1));
if(PrewittNum > PrewittThreshold)
newGrayPic(j,k)=255;
else
newGrayPic(j,k)=0;
end
end
end
figure,imshow(newGrayPic);
title(‘Prewitt算子的处理结果‘)
效果图如下：
方法四：Sobel算法
%Sobel 算子的实现：
clear;
sourcePic=imread(‘lena.jpg‘);
grayPic=mat2gray(sourcePic);
[m,n]=size(grayPic);
newGrayPic=grayPic;
sobelNum=0;
sobelThreshold=0.7;
for j=2:m-1
for k=2:n-1
sobelNum=abs(grayPic(j-1,k+1)+2*grayPic(j,k+1)+grayPic(j+1,k+1)-grayPic(j-1,k-1)-2*grayPic(j,k-1)-grayPic(j+1,k-1))+abs(grayPic(j-1,k-1)+2*grayPic(j-1,k)+grayPic(j-1,k+1)-grayPic(j+1,k-1)-2*grayPic(j+1,k)-grayPic(j+1,k+1));
if(sobelNum > sobelThreshold)
newGrayPic(j,k)=255;
else
newGrayPic(j,k)=0;
end
end
end
figure,imshow(newGrayPic);
title(‘Sobel算子的处理结果‘)
效果如下：
方法五：Roberts 算子的实现
%Roberts 算子的实现：
clear all;
clc;
sourcePic=imread(‘lena.jpg‘);
grayPic=mat2gray(sourcePic);
[m,n]=size(grayPic);
newGrayPic=grayPic;
robertsNum=0;
robertThreshold=0.2;
for j=1:m-1
for k=1:n-1
robertsNum = abs(grayPic(j,k)-grayPic(j+1,k+1)) + abs(grayPic(j+1,k)-grayPic(j,k+1));
if(robertsNum > robertThreshold)
newGrayPic(j,k)=255;
else
newGrayPic(j,k)=0;
end
end
end
figure,imshow(newGrayPic);
title(‘roberts算子的处理结果‘)
效果图：
参考：https://www.cnblogs.com/leegod/p/8109023.html
原文：https://www.cnblogs.com/Catherinezhilin/p/15088677.html