matlab三种图像二值化函数，详细，附代码（转）

代码链接：https://download.csdn.net/download/hupeng810/1511870

Ostu二值化：

thOTSU = graythresh(imag)
imagBWO = im2bw(imag, thOTSU)

function imagBW = otsu(imag)
function imagBW = otsu(imag)
% OTSU binarizes a gray scale image 'imag' into a binary image, with the
% noises removed.
% Input:
%   imag: the gray scale image, with black foreground(0), and white
%   background(255).
% Output:
%   imagBW: the binary image of the gray scale image 'imag', with Otsu
%   algorithm.

% Reference:
%   Nobuyuki Otsu. A Threshold Selection Method from Gray-Level Histograms.
%   IEEE Transactions on Systems, Man, and Cybernetics. 1979.SMC-9(1):62-66

imag = imag(:, :, 1);

[counts, x] = imhist(imag);  % counts are the histogram. x is the intensity level.
GradeI = length(x);   % the resolusion of the intensity. i.e. 256 for uint8.
varB = zeros(GradeI, 1);  % Between-class Variance of binarized image.

prob = counts ./ sum(counts);  % Probability distribution
meanT = 0;  % Total mean level of the picture
for i = 0 : (GradeI-1)
    meanT = meanT + i * prob(i+1);
end
varT = ((x-meanT).^2)' * prob; 
% Initialization
w0 = prob(1);   % Probability of the first class
miuK = 0;   % First-order cumulative moments of the histogram up to the kth level.
varB(1) = 0;
% Between-class variance calculation
for i = 1 : (GradeI-1)
    w0 = w0 + prob(i+1);
    miuK = miuK + i * prob(i+1);
    if (w0 == 0) || (w0 == 1)
        varB(i+1) = 0;
    else
        varB(i+1) = (meanT * w0 - miuK) .^ 2 / (w0 * (1-w0));
    end
end

maxvar = max(varB);
em = maxvar / varT  % Effective measure
index = find(varB == maxvar);
index = mean(index);
th = (index-1)/(GradeI-1)
imagBW = im2bw(imag, th);

% thOTSU = graythresh(imag)
% imagBWO = im2bw(imag, thOTSU);

Kittler二值化：

function imagBW = kittlerMet(imag)
% KITTLERMET binarizes a gray scale image 'imag' into a binary image
% Input:
%   imag: the gray scale image, with black foreground(0), and white
%   background(255).
% Output:
%   imagBW: the binary image of the gray scale image 'imag', with kittler's
%   minimum error thresholding algorithm.

% Reference:
%   J. Kittler and J. Illingworth. Minimum Error Thresholding. Pattern
%   Recognition. 1986. 19(1):41-47

MAXD = 100000;
imag = imag(:,:,1);
[counts, x] = imhist(imag);  % counts are the histogram. x is the intensity level.
GradeI = length(x);   % the resolusion of the intensity. i.e. 256 for uint8.
J_t = zeros(GradeI, 1);  % criterion function
prob = counts ./ sum(counts);  % Probability distribution
meanT = x' * prob;  % Total mean level of the picture
% Initialization
w0 = prob(1);   % Probability of the first class
miuK = 0;   % First-order cumulative moments of the histogram up to the kth level.
J_t(1) = MAXD; 
n = GradeI-1;
for i = 1 : n
    w0 = w0 + prob(i+1);
    miuK = miuK + i * prob(i+1);  % first-order cumulative moment
    if (w0 < eps) || (w0 > 1-eps)
        J_t(i+1) = MAXD;    % T = i
    else
        miu1 = miuK / w0;
        miu2 = (meanT-miuK) / (1-w0);
        var1 = (((0 : i)'-miu1).^2)' * prob(1 : i+1);
        var1 = var1 / w0;  % variance
        var2 = (((i+1 : n)'-miu2).^2)' * prob(i+2 : n+1);
        var2 = var2 / (1-w0);
        if var1 > eps && var2 > eps   % in case of var1=0 or var2 =0
            J_t(i+1) = 1+w0 * log(var1)+(1-w0) * log(var2)-2*w0*log(w0)-2*(1-w0)*log(1-w0);
        else
            J_t(i+1) = MAXD;
        end
    end
end
minJ = min(J_t);
index = find(J_t == minJ);
th = mean(index);
th = (th-1)/n
imagBW = im2bw(imag, th);

% figure, imshow(imagBW), title('kittler binary');

niblack二值化:

function imagBW = niblack(imag)
% NIBLACK binarizes a gray scale image 'imag' to a binary image, using
% Niblack algorithm. The noises of the gray scale image are removed.
%  Input:
%       imag: the gray scale image, with black foreground(0), and white
%       background(255).
%  Output:
%       imagBW: the binary image of the gray scale image 'imag', with
%       Niblack algorithm.

% Reference:
%   Wayne Niblack. An Introduction to Digital Image Processing. pp: 115.
%   1986. Prentice/Hall International. ISBN: 013 480674 3

tic;

k = -0.2;  % the first manual parameter
b = 80;   % the second manual parameter, about the width of the square neighborhood
choice = 1; % 1 for pixel-to-pixel computation, 2 for pixel averaging within the square neighborhood for fast computation.

imag = imag( :, :, 1);
[Hei, Wid] = size(imag);

imag = padarray(imag, [b b], 'symmetric', 'both');  % Pad image array 
Hei_pad = Hei + 2 * b;
Wid_pad = Wid + 2 * b;
imagBW = false(Hei_pad, Wid_pad);
switch choice
    case 1
        for i = 1+b : Hei+b
            for j = 1+b : Wid+b
                upR = i-floor(b/2-1/2);
                dnR = i+floor(b/2);
                lfC = j-floor(b/2-1/2);
                rtC = j+floor(b/2);
                m_ij = mean(mean(imag(upR : dnR, lfC : rtC)));
                sigma_squared = double(imag(upR : dnR, lfC : rtC)) - m_ij;
                sigma_squared = mean(mean(sigma_squared .^2));
                sigma = sqrt(sigma_squared);
                th_ij = m_ij + k * sigma;
                if double(imag(i,j)) > th_ij
                   imagBW(i,j) = 1;
                end
            end
        end
    case 2
        for i = 1+b : b : Hei+b
            for j = 1+b : b : Wid+b
                upR =  i-floor(b/2-1/2);
                dnR =  i+floor(b/2);
                lfC =  j-floor(b/2-1/2);
                rtC =  j+floor(b/2);
                m_ij = mean(mean(imag(upR : dnR, lfC : rtC)));
                sigma_squared = double(imag(upR : dnR, lfC : rtC)) - repmat(m_ij, (dnR-upR+1), (rtC-lfC+1));
                sigma_squared = sigma_squared .^ 2;
                sigma_squared = mean(mean(sigma_squared));
                sigma = sqrt(sigma_squared);
                th_ij = m_ij + k * sigma;
                imagBW(upR : dnR, lfC : rtC) = double(imag(upR : dnR, lfC : rtC)) > th_ij;
            end
        end
    otherwise
        display('Wrong Choice!');
end
imagBW = imagBW(1+b : Hei+b, 1+b : Wid+b);
% figure, imshow(imagBW), title('Binarized Image');
toc;