基于遗传算法的ostu阈值分割matlab代码

这是一个基于遗传算法的Otsu阈值分割的Matlab代码示例：

clc;
clear all;
close all;

% 读取图像
img = imread('lena512.bmp');
img = rgb2gray(img);

% 初始化遗传算法参数
popSize = 50;
numOfBits = 8;
numOfGenes = 1;
numOfIterations = 50;

% 像素值归一化到 [0,1] 区间
img = double(img) / 255;

% 计算直方图
histogram = histcounts(img, numOfBits) / numel(img);

% 初始化种群
population = zeros(popSize, numOfGenes);
for i = 1:popSize
    population(i,:) = randi([0, 2^numOfBits-1], [1, numOfGenes]);
end

% 迭代遗传算法
for iteration = 1:numOfIterations
    % 计算适应度函数
    fitness = zeros(popSize, 1);
    for i = 1:popSize
        threshold = population(i) / (2^numOfBits-1);
        foreground = img > threshold;
        background = img <= threshold;
        histogramForeground = histcounts(img(foreground), numOfBits) / sum(foreground(:));
        histogramBackground = histcounts(img(background), numOfBits) / sum(background(:));
        varianceWithinClass = sum((histogramForeground-mean(histogramForeground)).^2) + sum((histogramBackground-mean(histogramBackground)).^2);
        fitness(i) = varianceWithinClass;
    end
    
    % 选择操作
    [sortedFitness, indices] = sort(fitness);
    population = population(indices(1:popSize),:);
    
    % 交叉操作
    for i = 2:popSize
        if rand() < 0.8
            parent1 = population(i-1,:);
            parent2 = population(i,:);
            crossoverPoint = randi([1, numOfGenes-1]);
            child1 = [parent1(1:crossoverPoint), parent2(crossoverPoint+1:end)];
            child2 = [parent2(1:crossoverPoint), parent1(crossoverPoint+1:end)];
            population(i-1,:) = child1;
            population(i,:) = child2;
        end
    end
    
    % 变异操作
    for i = 1:popSize
        if rand() < 0.1
            gene = randi([0, 2^numOfBits-1], [1, numOfGenes]);
            population(i,:) = gene;
        end
    end
end

% 计算最优解
bestThreshold = population(1) / (2^numOfBits-1);

% 分割图像
foreground = img > bestThreshold;
background = img <= bestThreshold;

% 显示结果
figure();
subplot(1, 3, 1);
imshow(img);
title('原始图像');
subplot(1, 3, 2);
imshow(foreground);
title('前景');
subplot(1, 3, 3);
imshow(background);
title('背景');

这段代码首先读取一张灰度图像，将像素值归一化到 [0,1] 区间，计算直方图，并初始化遗传算法参数和种群。然后迭代遗传算法，计算每个个体的适应度函数，选择、交叉和变异操作，最终得到最优解作为阈值进行图像分割，显示分割结果。