Bilinear interpolation

Question

The following code is to enlarge pictures using bilinear interpolation. What can be modified in the function of slow_rescale to make it more efficient? I expect to modify it from the view of Principles of Computer Organization.

unsigned char *slow_rescale(unsigned char *src, int src_x, int src_y, int dest_x, int dest_y)
{
 double step_x,step_y;          // Step increase as per instructions above
 unsigned char R1,R2,R3,R4;     // Colours at the four neighbours
 unsigned char G1,G2,G3,G4;
 unsigned char B1,B2,B3,B4;
 double RT1, GT1, BT1;          // Interpolated colours at T1 and T2
 double RT2, GT2, BT2;
 unsigned char R,G,B;           // Final colour at a destination pixel
 unsigned char *dst;            // Destination image - must be allocated here! 
 int x,y;               // Coordinates on destination image
 double fx,fy;              // Corresponding coordinates on source image
 double dx,dy;              // Fractional component of source image    coordinates

 dst=(unsigned char *)calloc(dest_x*dest_y*3,sizeof(unsigned char));   // Allocate and clear   destination image
 if (!dst) return(NULL);                           // Unable to allocate image

 step_x=(double)(src_x-1)/(double)(dest_x-1);
 step_y=(double)(src_y-1)/(double)(dest_y-1);

 for (x=0;x<dest_x;x++)         // Loop over destination image
 for (y=0;y<dest_y;y++)
  {
fx=x*step_x;
fy=y*step_y;
dx=fx-(int)fx;
dy=fy-(int)fy;   
getPixel(src,floor(fx),floor(fy),src_x,&R1,&G1,&B1);    // get N1 colours
getPixel(src,ceil(fx),floor(fy),src_x,&R2,&G2,&B2); // get N2 colours
getPixel(src,floor(fx),ceil(fy),src_x,&R3,&G3,&B3); // get N3 colours
getPixel(src,ceil(fx),ceil(fy),src_x,&R4,&G4,&B4);  // get N4 colours
   // Interpolate to get T1 and T2 colours
   RT1=(dx*R2)+(1-dx)*R1;
   GT1=(dx*G2)+(1-dx)*G1;
   BT1=(dx*B2)+(1-dx)*B1;
   RT2=(dx*R4)+(1-dx)*R3;
   GT2=(dx*G4)+(1-dx)*G3;
   BT2=(dx*B4)+(1-dx)*B3;
   // Obtain final colour by interpolating between T1 and T2
   R=(unsigned char)((dy*RT2)+((1-dy)*RT1));
   G=(unsigned char)((dy*GT2)+((1-dy)*GT1));
   B=(unsigned char)((dy*BT2)+((1-dy)*BT1));
  // Store the final colour
  setPixel(dst,x,y,dest_x,R,G,B);
 } 
  return(dst);
}
void getPixel(unsigned char *image, int x, int y, int sx, unsigned char *R, unsigned char *G, unsigned char *B)
{
 // Get the colour at pixel x,y in the image and return it using the provided RGB pointers
 // Requires the image size along the x direction!
 *(R)=*(image+((x+(y*sx))*3)+0);
 *(G)=*(image+((x+(y*sx))*3)+1);
 *(B)=*(image+((x+(y*sx))*3)+2);
}

void setPixel(unsigned char *image, int x, int y, int sx, unsigned char R, unsigned char G, unsigned char B)
{
 // Set the colour of the pixel at x,y in the image to the specified R,G,B
 // Requires the image size along the x direction!
 *(image+((x+(y*sx))*3)+0)=R;
 *(image+((x+(y*sx))*3)+1)=G;
 *(image+((x+(y*sx))*3)+2)=B;
}

Answer 1

Please make everyone (including yourself) a favor :

• Declare your variable in the smallest possible scope. Also define them as your declare them if you can (and you usually can).
• Indent your code properly.

Then you'd get something like :

unsigned char *slow_rescale(unsigned char *src, int src_x, int src_y, int dest_x, int dest_y)
{
    unsigned char* dst=(unsigned char *)calloc(dest_x*dest_y*3,sizeof(unsigned char));   // Allocate and clear destination image
    if (!dst) return(NULL); // Unable to allocate image

    double step_x=(double)(src_x-1)/(double)(dest_x-1);
    double step_y=(double)(src_y-1)/(double)(dest_y-1);

    for (int x=0;x<dest_x;x++) // Loop over destination image
    {
        for (int y=0;y<dest_y;y++)
        {
            double fx=x*step_x;
            double fy=y*step_y;
            double dx=fx-(int)fx;
            double dy=fy-(int)fy;

            unsigned char R1,R2,R3,R4; // Colours at the four neighbours
            unsigned char G1,G2,G3,G4;
            unsigned char B1,B2,B3,B4;

            getPixel(src,floor(fx),floor(fy),src_x,&R1,&G1,&B1); // get N1 colours
            getPixel(src,ceil(fx),floor(fy),src_x,&R2,&G2,&B2); // get N2 colours
            getPixel(src,floor(fx),ceil(fy),src_x,&R3,&G3,&B3); // get N3 colours
            getPixel(src,ceil(fx),ceil(fy),src_x,&R4,&G4,&B4);  // get N4 colours
            // Interpolate to get T1 and T2 colours
            double RT1=(dx*R2)+(1-dx)*R1;
            double GT1=(dx*G2)+(1-dx)*G1;
            double BT1=(dx*B2)+(1-dx)*B1;
            double RT2=(dx*R4)+(1-dx)*R3;
            double GT2=(dx*G4)+(1-dx)*G3;
            double BT2=(dx*B4)+(1-dx)*B3;
            // Obtain final colour by interpolating between T1 and T2
            unsigned char R=(unsigned char)((dy*RT2)+((1-dy)*RT1));
            unsigned char G=(unsigned char)((dy*GT2)+((1-dy)*GT1));
            unsigned char B=(unsigned char)((dy*BT2)+((1-dy)*BT1));
            // Store the final colour
            setPixel(dst,x,y,dest_x,R,G,B);
        }
    }
    return(dst);
}
void getPixel(unsigned char *image, int x, int y, int sx, unsigned char *R, unsigned char *G, unsigned char *B)
{
    // Get the colour at pixel x,y in the image and return it using the provided RGB pointers
    // Requires the image size along the x direction!
    *(R)=*(image+((x+(y*sx))*3)+0);
    *(G)=*(image+((x+(y*sx))*3)+1);
    *(B)=*(image+((x+(y*sx))*3)+2);
}

void setPixel(unsigned char *image, int x, int y, int sx, unsigned char R, unsigned char G, unsigned char B)
{
    // Set the colour of the pixel at x,y in the image to the specified R,G,B
    // Requires the image size along the x direction!
    *(image+((x+(y*sx))*3)+0)=R;
    *(image+((x+(y*sx))*3)+1)=G;
    *(image+((x+(y*sx))*3)+2)=B;
}

Then, try to compute things once and only once when possible and you'll get something like :

unsigned char *slow_rescale(unsigned char *src, int src_x, int src_y, int dest_x, int dest_y)
{
    unsigned char* dst=(unsigned char *)calloc(dest_x*dest_y*3,sizeof(unsigned char));   // Allocate and clear destination image
    if (!dst) return(NULL); // Unable to allocate image

    double step_x=(double)(src_x-1)/(double)(dest_x-1);
    double step_y=(double)(src_y-1)/(double)(dest_y-1);

    for (int x=0;x<dest_x;x++) // Loop over destination image
    {
        double fx=x*step_x;
        double dx=fx-(int)fx;
        int ffx = floor(fx);
        int cfx = ceil(fx);
        for (int y=0;y<dest_y;y++)
        {
            double fy=y*step_y;
            double dy=fy-(int)fy;
            int ffy = floor(fy);
            int cfy = ceil(fy);

            unsigned char R1,R2,R3,R4,G1,G2,G3,G4,B1,B2,B3,B4;

            getPixel(src, (ffx + ffy*src_x)*3, &R1,&G1,&B1); // get N1 colours
            getPixel(src, (cfx + ffy*src_x)*3, &R2,&G2,&B2); // get N2 colours
            getPixel(src, (ffx + cfy*src_x)*3, &R3,&G3,&B3); // get N3 colours
            getPixel(src, (cfx + cfy*src_x)*3, &R4,&G4,&B4); // get N4 colours
            // Interpolate to get T1 and T2 colours
            double RT1 = dx*R2 + (1-dx)*R1;
            double GT1 = dx*G2 + (1-dx)*G1;
            double BT1 = dx*B2 + (1-dx)*B1;
            double RT2 = dx*R4 + (1-dx)*R3;
            double GT2 = dx*G4 + (1-dx)*G3;
            double BT2 = dx*B4 + (1-dx)*B3;
            // Obtain final colour by interpolating between T1 and T2
            unsigned char R = dy*RT2 + (1-dy)*RT1;
            unsigned char G = dy*GT2 + (1-dy)*GT1;
            unsigned char B = dy*BT2 + (1-dy)*BT1;
            // Store the final colour
            setPixel(dst, (x + y*dest_x)*3,R,G,B);
        }
    }
    return(dst);
}

inline void getPixel(unsigned char *image, int offset, unsigned char *R, unsigned char *G, unsigned char *B)
{
    *R=image[offset+0];
    *G=image[offset+1];
    *B=image[offset+2];
}

inline void setPixel(unsigned char *image, int offset, unsigned char R, unsigned char G, unsigned char B)
{
    image[offset+0]=R;
    image[offset+1]=G;
    image[offset+2]=B;
}

I'd expect this to be much much faster than your original code and to do roughtly the same thing (I have have introduced errors though).