convolution series naive

#include<iostream>
#include<cstdio>
#include<cstdlib>

#define Image_Height 20000
#define Image_Width  10000
#define Kernel_Dim 5  // convolutin kernel will be of width [2(r)+1]*[2(r)+1]
using namespace std;

int main()
{
    // Intializing image size and  intially seeding it with some randome values 0-255 (gray scale)

    int **image, **out_image;

    image = new int *[Image_Height];
    for(int i=0;i<Image_Height;i++)
    image[i] = new int[Image_Width];

    out_image = new int *[Image_Height];
    for(int i=0;i<Image_Height;i++)
    out_image[i] = new int[Image_Width];



    for(int i=0;i<Image_Height;i++)
    for(int j=0;j<Image_Width;j++)
    image[i][j]=rand()%256;


    // Intializing kernel let say guassian kernel of 5*5

    int kernel[Kernel_Dim][Kernel_Dim]={};
    int a[5]={1,4,6,4,1}; // to get gussian kernel by multiplying a*a we get gussian kernel of size 5*5

    for(int k=0;k<5;k++)
    for(int i=0;i<Kernel_Dim;i++)
    for(int j=0;j<Kernel_Dim;j++)
    kernel[i][j]=a[i]*a[j];


  int i,j,k,l,sum,r=2,count=0;
  // Main calulation done here multiplying image with kernel matrix
     // boundaries values skipped
  for(i=2;i<Image_Height-2;i++)
   {
       for(j=2;j<Image_Width-2;j++)
    {
          sum=0;
          for(k=-r;k<r;k++)
            {
                for(l=-r;l<r;l++)
                {
                    sum=sum+image[i+k][j+l]*kernel[k][l];

                }
            }
            out_image[i][j]=sum;
    }

   }









 return 0;
}