Algorithm 3: Image noise enhancement algorithm based on mean filtering

Input: image kernel_size

Output: enhanced_image

1:  functionenhance_noise_with_average_filter(image, kernel_size)

2:  height, width = image.shape

3:  Creates a blank image of the same size as the original image.

4:  Gets the height and width of the image.

5:  Performs noise enhancement on the image.

6:  for y = 0 to height − 1 do

7:   for x = 0 to width − 1 do

8:      Calculate the boundary coordinates of the filter. Assume that the coordinates of the computational boundary are $(x,y)$. $$\begin{array}{c}(x.y)=\left\{\begin{array}{c}\mathrm{top}=max(y-\lfloor N/2\rfloor ,0)\hfill \\ \mathrm{bottom}=min(y+\lfloor N/2\rfloor ,\mathrm{height}-1)\hfill \\ \mathrm{left}=max(x-\lfloor N/2\rfloor ,0)\hfill \\ \mathrm{right}=min(x+\lfloor N/2\rfloor ,\mathrm{width}-1)\hfill \end{array}\right.\end{array}$$

9:      The average value is calculated for the pixels within the filter. $$\begin{array}{c}\hfill {\mathrm{sum}}_{-}\mathrm{value}={\displaystyle \sum _{i=\mathrm{top}}^{\mathrm{bottom}}}{\displaystyle \sum _{j=\mathrm{left}}^{\mathrm{right}}}\mathrm{image}(x+i-\lfloor N/2\rfloor ,y+j-\lfloor N/2\rfloor )\end{array}$$

10:      The average value is used as the enhanced pixel value. $$\begin{array}{c}\hfill \mathrm{average}=\frac{\mathrm{sum}\_\mathrm{value}}{N\times N}\end{array}$$

11:   end for

12:  end for

13:  return enhanced_image

14:  end function