Android bitmap operations using ndk - image-processing

Question

Currently I'm developing an Android application that involves some image processing. After some research I have made I found that is better to use Android NDK for bitmap manipulation for a good performance. So, I found some basic examples like this one:
static void myFunction(AndroidBitmapInfo* info, void* pixels){
int xx, yy, red, green, blue;
uint32_t* line;
for(yy = 0; yy < info->height; yy++){
line = (uint32_t*)pixels;
for(xx =0; xx < info->width; xx++){
//extract the RGB values from the pixel
blue = (int) ((line[xx] & 0x00FF0000) >> 16);
green = (int)((line[xx] & 0x0000FF00) >> 8);
red = (int) (line[xx] & 0x00000FF );
//change the RGB values
// set the new pixel back in
line[xx] =
((blue << 16) & 0x00FF0000) |
((green << 8) & 0x0000FF00) |
(red & 0x000000FF);
}
pixels = (char*)pixels + info->stride;
}
}
I used this code and it works very well for basic operations, but I want to make a more complex one, like a filter, where I need to access the above and below pixels from the current pixel. To be more specific, I'll give you an example: for dilation and erosion operations we move through pixels and we verify if the pixels from north west, north, north east, west, east, south west, south and south east (for 8 neighbors structure element) are object pixels. What I need to know is how can I access the values of the north and south pixels using the above code.
I'm not very familiarized with image processing using C (pointers etc.).
Thanks!

Answer 1

I've edit your function a little, basically to get the pixel position in the array the formula is:
position = y*width+x
static void myFunction(AndroidBitmapInfo* info, void* pixels){
int xx, yy, red, green, blue;
uint32_t* px = pixels;
for(yy = 0; yy < info->height; yy++){
for(xx =0; xx < info->width; xx++){
int position = yy*info->width+xx;//this formula gives you the address of pixel with coordinates (xx; yy) in 'px'
//extract the RGB values from the pixel
blue = (int) ((line[position] & 0x00FF0000) >> 16);
green = (int)((line[position] & 0x0000FF00) >> 8);
red = (int) (line[position] & 0x00000FF );
//change the RGB values
// set the new pixel back in
line[position] =
((blue << 16) & 0x00FF0000) |
((green << 8) & 0x0000FF00) |
(red & 0x000000FF);
//so the position of the south pixel is (yy+1)*info->width+xx
//and the one of the north is (yy-1)*info->width+xx
//the left yy*info->width+xx-1
//the right yy*info->width+xx+1
}
}
assuming you want to read/edit a pixel with coordinates x,y
you must check if 0 <= y < height and if 0 <= x < width , otherwise you may access unexisting pixels and you will have a memory access error