直方图均衡化matlab但是不要使用histeq，自己编写循环计算

直方图均衡化是一种常见的图像增强技术，其目的是改善图像对比度，使其细节更容易识别。在MATLAB中，可以手动编写循环来实现直方图均衡化，避免直接使用`histeq`函数。这里是一个基本的步骤：

1. **读取图像并计算直方图**：

img = imread('your_image.jpg'); % 替换为你要处理的图片路径
height = size(img, 1);
width = size(img, 2);
num_channels = size(img, 3);

% 如果是三通道图像，计算每个通道的直方图
if num_channels == 3
    histSize = 256; % 直方图大小
    channelHistograms = zeros(histSize, 3);
else
    histSize = 256;
    channelHistograms = zeros(histSize, 1);
end

for y = 1:height
    for x = 1:width
        channelValues = img(y, x, :); % 获取单个像素或通道值
        
        % 更新每个通道的直方图
        if num_channels == 3
            channelHistograms(:, 1) = channelHistograms(:, 1) + histcounts(channelValues(1), 1:histSize);
            channelHistograms(:, 2) = channelHistograms(:, 2) + histcounts(channelValues(2), 1:histSize);
            channelHistograms(:, 3) = channelHistograms(:, 3) + histcounts(channelValues(3), 1:histSize);
        else
            channelHistograms(:, 1) = channelHistograms(:, 1) + histcounts(channelValues, 1:histSize);
        end
    end
end

2. **计算累计分布函数（CDF）**：

cumulativeDist = cumsum(channelHistograms, 1);

3. **计算等概率分布**：

equiDist = linspace(1, histSize, histSize);

4. **归一化像素**：

equiImage = zeros(height, width, num_channels);
for y = 1:height
    for x = 1:width
        for c = 1:num_channels
            if num_channels == 3
                pixelValue = img(y, x, c);
                eqPixelValue = interp1(equiDist(c), cumulativeDist(c), pixelValue, 'linear', 'extrap');
            else
                eqPixelValue = interp1(equiDist, cumulativeDist, img(y, x), 'linear', 'extrap');
            end
            
            equiImage(y, x, c) = round(eqPixelValue);
        end
    end
end

现在，`equiImage`就是经过了手工直方图均衡化的图像。