【C++】【图像处理】均值滤波 and 高斯滤波 and 中值滤波 （低通滤波器）and Sobel算子边缘提取算法解析（以.raw格式的图像为基础进行图像处理、gray levels：256）

 1 void meanFilter(BYTE* image, int width, int height, BYTE* outImg)
 2 {
 3     //均值滤波
 4     int smth[9];
 5     int i, j, m, n;
 6     BYTE block[9];
 7     
 8     // 高斯卷积核初始化
 9     smth[0] = 1, smth[1] = 2, smth[2] = 1,
10     smth[3] = 2, smth[4] = 4, smth[5] = 2,
11     smth[6] = 1, smth[7] = 2, smth[8] = 1;
12 
13     int value;
14     for (i = 0;i < 9;i++) //初始化均值卷积核
15         smth[i] = 1;
16 
17     for (i = 0;i < height;i++) 
18         for (j = 0;j < width;j++) {
19             //将输出图像边缘的像素值设为 0
20             if (i == 0 || j == 0 || i == height - 1 || j == width - 1) 
21                 outImg[i * width + j] = 0;
22 
23             //提取以当前像素为中心的 3x3 区域的像素值，然后利用卷积操作计算这个区域的均值，最后将计算得到的均值作为输出图像中对应位置的像素值
24             else {
25                 for (m = -1;m < 2;m++) 
26                     for (n = -1;n < 2;n++) 
27                         block[(m + 1) * 3 + n + 1] = image[(i + m) * width + j + n];
28                 value = convolution(smth, block);
29                 outImg[i * width + j] = BYTE(value / 9.0f); // 高斯为 value / 16.0f
30             }
31         }
32 
33 }
34 
35 
36 int convolution(int* operatr, BYTE* block)
37 {
38     int value;
39     int i, j;
40     value = 0;
41     //卷积运算
42     //遍历 3x3 区域内的每个像素，并根据卷积核的权重计算出卷积结果
43     for (i = 0;i < 3;i++) 
44         for (j = 0;j < 3;j++) 
45             value += operatr[i * 3 + j] * block[i * 3 + j];
46     /*
47     1 1 1
48     1 1 1
49     1 1 1
50 
51     0 0 0
52     0 x y
53     0 j k
54     */
55 
56     return value;
57 
58 }  

 1 void midFindFiltering(BYTE* image, int width, int height, BYTE* outImg)
 2 {
 3     //中值滤波
 4     int i, j, m, n;
 5     BYTE block[9];
 6 
 7     int value;
 8 
 9     for (i = 0;i < height;i++) 
10         for (j = 0;j < width;j++) {
11             if (i == 0 || j == 0 || i == height - 1 || j == width - 1)
12                 outImg[i * width + j] = 0;
13             else {
14                 for (m = -1;m < 2;m++)
15                     for (n = -1;n < 2;n++)
16                         block[(m + 1) * 3 + n + 1] = image[(i + m) * width + j + n];
17             }
18             value = MidValueFind(sizeof(block), block);
19             outImg[i * width + j] = value;
20         }
21 }

int MidValueFind(int num, BYTE* d)
{
    int value;

    int i, j;
    int temp;
    //冒泡排序
    for (i = 0; i < num - 1; i++)
        for (j = i + 1; j < num; j++)
        {
            if (d[i] < d[j])
            {
                temp = d[i];
                d[i] = d[j];
                d[j] = temp;
            }
        }

    return d[num / 2];

}

 1 void sobel(BYTE* window, int wid, int hei, int* sob_x, int* sob_y)
 2 {
 3     int so_x[9];//horizontal 水平边缘提取卷积核初始化
 4     so_x[0] = -1, so_x[1] = 0, so_x[2] = 1,
 5     so_x[3] = -2, so_x[4] = 0, so_x[5] = 2,
 6     so_x[6] = -1, so_x[7] = 0, so_x[8] = 1;
 7 
 8     int so_y[9];//vertical   垂直边缘提取卷积核初始化
 9     so_y[0] = -1,   so_y[1] = -2,   so_y[2] = -1,
10     so_y[3] = 0,    so_y[4] = 0,    so_y[5] = 0,
11     so_y[6] = 1,    so_y[7] = 2,    so_y[8] = 1;
12 
13     int i, j, m, n;
14     BYTE block[9];
15     int value;
16 
17     //求图像边缘
18     for (i = 0;i < hei;i++) 
19         for (j = 0;j < wid;j++) {
20             if (i == 0 || j == 0 || i == hei - 1 || j == wid - 1) {
21                 sob_x[i * wid + j] = sob_y[i * wid + j] = 0; // 输出图像边缘归零
22             }
23             else {
24                 for (m = -1;m < 2;m++)
25                     for (n = -1;n < 2;n++)
26                         block[(m + 1) * 3 + n + 1] = window[(i + m) * wid + j + n];
27             }
28             sob_x[i * wid + j] = convolution(so_x, block);
29             sob_y[i * wid + j] = convolution(so_y, block);
30 
31         }
32 }