Jump to content

  • Log In with Google      Sign In   
  • Create Account


Splitting an image into tiles


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
3 replies to this topic

#1 george7378   Members   -  Reputation: 1132

Like
0Likes
Like

Posted 05 May 2014 - 06:12 PM

Hi!

 

Let's say I have an image and I want to loop through the pixels. I know how to do this in scan-line form:

 

for (x = 0; x < w; x++)
{
for (y = 0; y < h; y++)
{
currentPixel = image[y*w + x];
}
}

 

How would I do the same thing if I was splitting my image into square tiles?

 

Thanks :)



Sponsor:

#2 Bacterius   Crossbones+   -  Reputation: 8134

Like
2Likes
Like

Posted 05 May 2014 - 06:25 PM

Something like this? Pseudocode, of course, four nested loops is unsightly and can probably be refactored depending on your code, e.g. foreach tile process tile.

for (tileY = 0; tileY < h / tileH; ++tileY)
    for (tileX = 0; tileX < w / tileW; ++tileX)
        for (y = tileY * tileH; y < (tileY + 1) * tileH; ++y)
            for (x = tileX * tileW; x < (tileX + 1) * tileW; ++x)
            {
                // do stuff on pixel (x, y)
            }

Obviously the image needs to split properly into tiles of the specified dimension (tileW, tileH), otherwise you'll need to handle the leftover pixels somehow.


The slowsort algorithm is a perfect illustration of the multiply and surrender paradigm, which is perhaps the single most important paradigm in the development of reluctant algorithms. The basic multiply and surrender strategy consists in replacing the problem at hand by two or more subproblems, each slightly simpler than the original, and continue multiplying subproblems and subsubproblems recursively in this fashion as long as possible. At some point the subproblems will all become so simple that their solution can no longer be postponed, and we will have to surrender. Experience shows that, in most cases, by the time this point is reached the total work will be substantially higher than what could have been wasted by a more direct approach.

 

- Pessimal Algorithms and Simplexity Analysis


#3 Ravyne   Crossbones+   -  Reputation: 6728

Like
3Likes
Like

Posted 05 May 2014 - 06:40 PM

So, first of all, fix your loops -- having y inside x is terrible for your cache performance.

 

The general answer to tiling the image is to wrap your y-x pixel loop in another y-x tile loop. Divide the height and width of the whole image by the height and width of your tiles, respectively, to get the number of iterations in your y-x tile loops. You'll need to take special care of images that don't divide evenly into perfect tiles.


Edited by Ravyne, 05 May 2014 - 06:44 PM.


#4 ferrous   Members   -  Reputation: 1535

Like
0Likes
Like

Posted 06 May 2014 - 04:10 PM

Reminds me that one of the things we wrote as interns back in the old Game Boy Advance days, was a tool to check the image for repeated tiles.  Is one tile nearly the same as another tile, what about if the tile was rotated by some 90 degree increment, would it match then, etc.






Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS