自适应加权中值滤波算法matlab

自适应加权中值滤波算法（Adaptive Weighted Median Filter）是一种常用的非线性滤波算法，可用于去除高斯噪声、脉冲噪声等。
Matlab代码实现如下：

function [output_img] = AWMedianFilter(input_img, max_window_size)
% 自适应加权中值滤波
% input_img: 输入图像
% max_window_size: 最大窗口尺寸
    [m, n] = size(input_img);
    output_img = zeros(m, n); % 输出图像
    padding_size = floor(max_window_size/2);
    for i = 1:m
        for j = 1:n
            cur_pixel = input_img(i,j); % 当前像素值
            window_size = 3; % 窗口大小从3开始
            while window_size <= max_window_size
                padding = floor(window_size/2);
                window = input_img(max(i-padding,1):min(i+padding,m), ...
                    max(j-padding,1):min(j+padding,n)); % 提取窗口
                if cur_pixel > min(window(:)) && cur_pixel < max(window(:))
                    % 计算加权中值
                    sorted_window = sort(window(:));
                    center_pixel = sorted_window(padding+1);
                    left_window = sorted_window(1:padding);
                    right_window = sorted_window(padding+2:end);
                    left_weight = sum(center_pixel-left_window) / padding;
                    right_weight = sum(right_window-center_pixel) / padding;
                    output_img(i,j) = center_pixel + 0.5 * (left_weight - right_weight);
                    break; % 找到加权中值，跳出循环
                else
                    window_size = window_size + 2; % 窗口大小加2
                end
            end
            if window_size > max_window_size % 最大窗口尺寸内无法找到加权中值，取原像素值
                output_img(i,j) = cur_pixel;
            end
        end
    end
    output_img = uint8(output_img); % 转为uint8类型
end

调用方法：

input_img = imread('lena.png');
output_img = AWMedianFilter(input_img, 7);
imshow(input_img), title('原图');
figure, imshow(output_img), title('滤波后的图像');