matlab实现ostu算法程序

下面是一个简单的 Matlab 实现 Ostsu 算法的程序：

function [threshold, level] = ostu(I)
% OSTU algorithm implementation
% I: input grayscale image
% threshold: threshold value
% level: binary image with 0 and 1 values

% Calculate histogram
[counts, ~] = imhist(I);

% Normalize histogram
p = counts / sum(counts);

% Initialize variables
level = 0:255;
sigma_b = zeros(256,1);

% Compute sigma_b for each level
for t=1:256
    q1 = sum(p(1:t)); % probability of class 1
    q2 = 1 - q1; % probability of class 2
    if (q1 == 0 || q2 == 0)
        continue;
    end
    mu1 = sum(level(1:t).*p(1:t)) / q1; % mean of class 1
    mu2 = sum(level(t+1:end).*p(t+1:end)) / q2; % mean of class 2
    sigma_b(t) = q1 * q2 * (mu1 - mu2)^2; % compute sigma_b
end

% Find the optimum threshold
[~,threshold] = max(sigma_b);

% Create binary image
level(I < threshold) = 0;
level(I >= threshold) = 1;
end

使用方法：

% Load image
I = imread('image.jpg');
% Call ostu function
[threshold, level] = ostu(I);
% Display binary image
imshow(level);

这个程序将加载名为 image.jpg 的图像，执行 Ostsu 算法，并显示二进制图像。