CS50 Pset4 过滤器（不太舒服）模糊功能算法问题

Question

info about this task

当我尝试实现模糊功能时，它在图片上对我来说效果很好，但 check50（cs50 测试程序）对我的输出发出警告。

这是我的代码

void blur(int height, int width, RGBTRIPLE image[height][width])
{
    float average_red = 0;
    float average_green = 0;
    float average_blue = 0;
    float count = 0;
    for (int i = 0; i < height; i++)
    {
        for (int j = 0; j < width; j++)
        {
            // reset values 
            average_red = 0;
            average_blue = 0;
            average_green = 0;
            count = 0;

            // look for around a pixel 3x3 box
            // column
            for (int k = i - 1; k < i + 2; k++)
            {
                // row
                for (int l = j - 1; l < j + 2; l++)
                {
                    // if pixel on the top
                    if (k == -1)
                    {
                        // it skips a column because it is out of the border
                        break;
                    }

                    // if pixel is on the left side
                    if (l == -1)
                    {
                        // skips a row otherwise it is out of the border
                        continue;
                    }

                    // if pixel passes the bottom
                    if (k >= height)
                    {
                        break;
                    }

                    // if pixels passes the right side
                    if (l >= width)
                    {
                        continue;
                    }
                    
                    // everything else
                    else
                    {
                        average_red += image[k][l].rgbtRed;
                        average_green += image[k][l].rgbtGreen;
                        average_blue += image[k][l].rgbtBlue;
                        count++;
                    }
                }   
            }

            average_red /= count;
            average_green /= count;
            average_blue /= count;
            image[i][j].rgbtRed  = round(average_red);
            image[i][j].rgbtGreen = round(average_green);
            image[i][j].rgbtBlue = round(average_blue);

        }
    }
    return;
}

expected output vs my output is here（模糊函数在下面输出）

它在拐角处工作正常，但其他像素值非常接近正确的输出，但不一样。 任何帮助表示赞赏

Answer 1

代码需要考虑几个因素。建议：

如果结果行号是 (height-1) 则不计算该像素

如果生成的列号为 (width-1)，则不计算该像素

因此，对于周围 8 个像素中的每一个，应用上述两个标准。

Answer 2

我自己的问题解决了，还是谢谢你的帮助

问题在于我的总和计算。它将结果保留在 image[i][j] 中，然后再次使用该值。 我将图像值复制到一个新值并用于此。这是我的代码

void blur(int height, int width, RGBTRIPLE image[height][width])
{
    // copy original value to a new value to keep changing values
    RGBTRIPLE copy[height][width];
    for (int i = 0; i < height; i++)
    {
        for (int j = 0; j < width; j++)
        {
            copy[i][j] = image[i][j];
        }
    }

    float average_red = 0.0f;
    float average_green = 0.0f;
    float average_blue = 0.0f;
    float count = 0;

    for (int i = 0; i < height; i++)
    {
        for (int j = 0; j < width; j++)
        {
            // reset values 
            average_red = 0.0f;
            average_blue = 0.0f;
            average_green = 0.0f;
            count = 0;

            // look for around a pixel 3x3 box
            // column
            for (int k = i - 1; k < i + 2; k++)
            {
                // row
                for (int l = j - 1; l < j + 2; l++)
                {
                    // if pixel on the top
                    if (k == -1)
                    {
                        // it skips a column because it is out of the border
                        break;
                    }

                    // if pixel is on the left side
                    else if (l == -1)
                    {
                        // skips a row otherwise it is out of the border
                        continue;
                    }

                    // if pixel passes the bottom
                    else if (k >= height)
                    {
                        break;
                    }

                    // if pixels passes the right side
                    else if (l >= width)
                    {
                        continue;
                    }
                    
                    // everything else
                    else
                    {
                        average_red += copy[k][l].rgbtRed;
                        average_green += copy[k][l].rgbtGreen;
                        average_blue += copy[k][l].rgbtBlue;
                        count++;
                    }
                }   
            }

            average_red /= count;
            average_green /= count;
            average_blue /= count;
            image[i][j].rgbtRed  = round(average_red);
            image[i][j].rgbtGreen = round(average_green);
            image[i][j].rgbtBlue = round(average_blue);

        }
    }
    return;
}