动态二值化代码(Niblack)

        本篇文章主要给出Niblack动态二值化代码，鉴于自身水平所限，如有错误，欢迎批评指正。（欢迎进Q群交流：874653199）

        二值化结果：

matlab代码：

%Niblack算法 
clc;
clear;

rawImg=imread('E:\2_光学测量\21_格雷码重建\3_数据\graycode\01\15.bmp');

srcImg = rawImg;

if(length(size(srcImg))>2)
    
    srcImg=rgb2gray(srcImg);
    
end
 
w03 =35;%   
max03 = 0;   
min03 = 0;   
[m,n] = size(srcImg);   
T03 = zeros(m ,n );   
  
%
for i = (w03 + 1):(m - w03)   
    for j = (w03 + 1):(n - w03)      
        sum = 0;
        for k = -w03:w03   
            for l = -w03:w03   
                sum = sum + uint32(srcImg(i + k,j + l));
            end   
        end   
        average = double(sum) /((2*w03+1)*(2*w03+1));
        s = 0;
        for k = -w03:w03   
            for l = -w03:w03   
                s = s +   (uint32(srcImg(i + k,j + l)) - average)*(uint32(srcImg(i + k,j + l)) - average);
            end   
        end   
        s= sqrt(double(s)/((2*w03+1)*(2*w03+1)));
        
        T03(i,j) = average + 0.2*s;
    end   
end   
for i =  1:m 
    for j = 1:n 
        if srcImg(i,j) > T03(i,j)   
            srcImg(i,j) = uint8(255);   
        else  
            srcImg(i,j) = uint8(0);   
        end   
    end   
end   

t=graythresh(srcImg);
bwImages =imbinarize(rawImg,t);%OTSU二值化;

subplot(1,3,1);imshow(rawImg);  title('原图像')
subplot(1,3,2);imshow(srcImg);  title('Niblack算法')
subplot(1,3,3);imshow(bwImages);  title('OTSU算法')