给图像分别添加高斯噪声、椒盐噪声和均匀分布的噪声， 然后分别使用空间滤波器进行图像复原，要求使用均值滤波 器中至少四种滤波器进行处理，并比较各滤波器的效果。C++

以下是使用C++进行图像噪声添加和空间滤波的示例代码：

#include <opencv2/opencv.hpp>
#include <iostream>
#include <cstdlib>
#include <ctime>

using namespace cv;
using namespace std;

// 添加高斯噪声
void addGaussianNoise(Mat& img, double mean, double stddev)
{
    Mat noise = Mat::zeros(img.size(), CV_8UC1);
    randn(noise, mean, stddev); // 生成高斯噪声矩阵
    img += noise; // 像素值加上噪声值
}

// 添加椒盐噪声
void addSaltPepperNoise(Mat& img, double ratio)
{
    int num = img.rows * img.cols * ratio;
    for (int i = 0; i < num; i++)
    {
        int x = rand() % img.cols;
        int y = rand() % img.rows;
        if (rand() % 2 == 0)
            img.at<uchar>(y, x) = 0; // 盐噪声
        else
            img.at<uchar>(y, x) = 255; // 椒噪声
    }
}

// 添加均匀分布噪声
void addUniformNoise(Mat& img, double a, double b)
{
    Mat noise = Mat::zeros(img.size(), CV_8UC1);
    randu(noise, Scalar(a), Scalar(b)); // 生成均匀分布噪声矩阵
    img += noise; // 像素值加上噪声值
}

// 均值滤波器
Mat meanFilter(const Mat& img, int ksize)
{
    Mat result = Mat::zeros(img.size(), CV_8UC1);
    int border = ksize / 2;
    for (int i = border; i < img.rows - border; i++)
    {
        for (int j = border; j < img.cols - border; j++)
        {
            int sum = 0;
            for (int y = -border; y <= border; y++)
            {
                for (int x = -border; x <= border; x++)
                {
                    sum += img.at<uchar>(i + y, j + x);
                }
            }
            result.at<uchar>(i, j) = sum / (ksize * ksize);
        }
    }
    return result;
}

// 加权均值滤波器
Mat weightedMeanFilter(const Mat& img, int ksize)
{
    Mat result = Mat::zeros(img.size(), CV_8UC1);
    int border = ksize / 2;
    double weight = 1.0 / (ksize * ksize);
    double centerWeight = (ksize * ksize - 1) * weight;
    for (int i = border; i < img.rows - border; i++)
    {
        for (int j = border; j < img.cols - border; j++)
        {
            double sum = 0;
            for (int y = -border; y <= border; y++)
            {
                for (int x = -border; x <= border; x++)
                {
                    if (y == 0 && x == 0)
                        sum += centerWeight * img.at<uchar>(i + y, j + x);
                    else
                        sum += weight * img.at<uchar>(i + y, j + x);
                }
            }
            result.at<uchar>(i, j) = sum;
        }
    }
    return result;
}

// 中值滤波器
Mat medianFilter(const Mat& img, int ksize)
{
    Mat result = Mat::zeros(img.size(), CV_8UC1);
    int border = ksize / 2;
    vector<uchar> values(ksize * ksize);
    for (int i = border; i < img.rows - border; i++)
    {
        for (int j = border; j < img.cols - border; j++)
        {
            int index = 0;
            for (int y = -border; y <= border; y++)
            {
                for (int x = -border; x <= border; x++)
                {
                    values[index++] = img.at<uchar>(i + y, j + x);
                }
            }
            sort(values.begin(), values.end());
            result.at<uchar>(i, j) = values[ksize * ksize / 2];
        }
    }
    return result;
}

// 自适应加权平均滤波器
Mat adaptiveWeightedMeanFilter(const Mat& img, int ksize, double delta)
{
    Mat result = Mat::zeros(img.size(), CV_8UC1);
    int border = ksize / 2;
    double weight = 1.0 / (ksize * ksize);
    for (int i = border; i < img.rows - border; i++)
    {
        for (int j = border; j < img.cols - border; j++)
        {
            double sum = 0;
            double center = img.at<uchar>(i, j);
            for (int y = -border; y <= border; y++)
            {
                for (int x = -border; x <= border; x++)
                {
                    double diff = abs(img.at<uchar>(i + y, j + x) - center);
                    double w = diff <= delta ? weight / (1 + diff * diff / (delta * delta)) : 0;
                    sum += w * img.at<uchar>(i + y, j + x);
                }
            }
            result.at<uchar>(i, j) = sum;
        }
    }
    return result;
}

int main()
{
    srand(time(NULL));

    Mat img = imread("lena.jpg", IMREAD_GRAYSCALE);
    if (img.empty())
    {
        cout << "Failed to read image" << endl;
        return -1;
    }

    // 添加高斯噪声
    Mat img1 = img.clone();
    addGaussianNoise(img1, 0, 10);

    // 添加椒盐噪声
    Mat img2 = img.clone();
    addSaltPepperNoise(img2, 0.1);

    // 添加均匀分布噪声
    Mat img3 = img.clone();
    addUniformNoise(img3, -50, 50);

    // 均值滤波器
    Mat mean1 = meanFilter(img1, 3);
    Mat mean2 = meanFilter(img1, 5);
    Mat mean3 = meanFilter(img1, 7);
    Mat mean4 = meanFilter(img1, 9);

    // 加权均值滤波器
    Mat weightedMean1 = weightedMeanFilter(img1, 3);
    Mat weightedMean2 = weightedMeanFilter(img1, 5);
    Mat weightedMean3 = weightedMeanFilter(img1, 7);
    Mat weightedMean4 = weightedMeanFilter(img1, 9);

    // 中值滤波器
    Mat median1 = medianFilter(img1, 3);
    Mat median2 = medianFilter(img1, 5);
    Mat median3 = medianFilter(img1, 7);
    Mat median4 = medianFilter(img1, 9);

    // 自适应加权平均滤波器
    Mat adaptiveWeightedMean1 = adaptiveWeightedMeanFilter(img1, 3, 10);
    Mat adaptiveWeightedMean2 = adaptiveWeightedMeanFilter(img1, 5, 10);
    Mat adaptiveWeightedMean3 = adaptiveWeightedMeanFilter(img1, 7, 10);
    Mat adaptiveWeightedMean4 = adaptiveWeightedMeanFilter(img1, 9, 10);

    // 显示结果
    namedWindow("Original", WINDOW_NORMAL);
    namedWindow("Gaussian Noise", WINDOW_NORMAL);
    namedWindow("Mean Filter 3x3", WINDOW_NORMAL);
    namedWindow("Mean Filter 5x5", WINDOW_NORMAL);
    namedWindow("Mean Filter 7x7", WINDOW_NORMAL);
    namedWindow("Mean Filter 9x9", WINDOW_NORMAL);
    namedWindow("Weighted Mean Filter 3x3", WINDOW_NORMAL);
    namedWindow("Weighted Mean Filter 5x5", WINDOW_NORMAL);
    namedWindow("Weighted Mean Filter 7x7", WINDOW_NORMAL);
    namedWindow("Weighted Mean Filter 9x9", WINDOW_NORMAL);
    namedWindow("Median Filter 3x3", WINDOW_NORMAL);
    namedWindow("Median Filter 5x5", WINDOW_NORMAL);
    namedWindow("Median Filter 7x7", WINDOW_NORMAL);
    namedWindow("Median Filter 9x9", WINDOW_NORMAL);
    namedWindow("Adaptive Weighted Mean Filter 3x3", WINDOW_NORMAL);
    namedWindow("Adaptive Weighted Mean Filter 5x5", WINDOW_NORMAL);
    namedWindow("Adaptive Weighted Mean Filter 7x7", WINDOW_NORMAL);
    namedWindow("Adaptive Weighted Mean Filter 9x9", WINDOW_NORMAL);

    imshow("Original", img);
    imshow("Gaussian Noise", img1);
    imshow("Mean Filter 3x3", mean1);
    imshow("Mean Filter 5x5", mean2);
    imshow("Mean Filter 7x7", mean3);
    imshow("Mean Filter 9x9", mean4);
    imshow("Weighted Mean Filter 3x3", weightedMean1);
    imshow("Weighted Mean Filter 5x5", weightedMean2);
    imshow("Weighted Mean Filter 7x7", weightedMean3);
    imshow("Weighted Mean Filter 9x9", weightedMean4);
    imshow("Median Filter 3x3", median1);
    imshow("Median Filter 5x5", median2);
    imshow("Median Filter 7x7", median3);
    imshow("Median Filter 9x9", median4);
    imshow("Adaptive Weighted Mean Filter 3x3", adaptiveWeightedMean1);
    imshow("Adaptive Weighted Mean Filter 5x5", adaptiveWeightedMean2);
    imshow("Adaptive Weighted Mean Filter 7x7", adaptiveWeightedMean3);
    imshow("Adaptive Weighted Mean Filter 9x9", adaptiveWeightedMean4);

    waitKey(0);
    return 0;
}

其中，`addGaussianNoise`、`addSaltPepperNoise`和`addUniformNoise`分别用于添加高斯噪声、椒盐噪声和均匀分布噪声；`meanFilter`、`weightedMeanFilter`、`medianFilter`和`adaptiveWeightedMeanFilter`分别用于实现均值滤波器、加权均值滤波器、中值滤波器和自适应加权平均滤波器。在`main`函数中，首先读入`lena.jpg`图像，然后依次对其添加高斯噪声、椒盐噪声和均匀分布噪声，并使用均值滤波器中的四种滤波器、加权均值滤波器、中值滤波器和自适应加权平均滤波器进行处理，最后将处理结果显示出来。