Flood fill is an algorithm mainly used to determine a bounded area connected to a given node in a multi-dimensional array. It is a close resemblance to the bucket tool in paint programs.

The most approached implementation of the algorithm is a stack-based recursive function, and that’s what we’re gonna talk about next.

How does it work?

The problem is pretty simple and usually follows these steps:

  1. Take the position of the starting point.
  2. Decide wether you want to go in 4 directions (N, S, W, E) or 8 directions (N, S, W, E, NW, NE, SW, SE).
  3. Choose a replacement color and a target color.
  4. Travel in those directions.
  5. If the tile you land on is a target, reaplce it with the chosen color.
  6. Repeat 4 and 5 until you’ve been everywhere within the boundaries.

Let’s take the following array as an example:

alt text

The red square is the starting point and the gray squares are the so called walls.

For further details, here’s a piece of code describing the function:

int wall = -1;

void flood_fill(int pos_x, int pos_y, int target_color, int color)
{

   if(a[pos_x][pos_y] == wall || a[pos_x][pos_y] == color) // if there is no wall or if i haven't been there
      return;                                              // already go back

   if(a[pos_x][pos_y] != target_color) // if it's not color go back
      return;

   a[pos_x][pos_y] = color; // mark the point so that I know if I passed through it. 

   flood_fill(pos_x + 1, pos_y, color);  // then i can either go south
   flood_fill(pos_x - 1, pos_y, color);  // or north
   flood_fill(pos_x, pos_y + 1, color);  // or east
   flood_fill(pos_x, pos_y - 1, color);  // or west

   return;

}

As seen above, my starting point is (4,4). After calling the function for the start coordinates x = 4 and y = 4, I can start checking if there is no wall or color on the spot. If that is valid i mark the spot with one “color” and start checking the other adiacent squares.

Going south we will get to point (5,4) and the function runs again.

Excercise problem

I always considered that solving a (or more) problem/s using a newly learned algorithm is the best way to fully understand the concept.

So here’s one:

Statement:

In a bidimensional array you are given n number of “islands”. Try to find the largest area of an island and the corresponding island number. 0 marks water and any other x between 1 and n marks one square from the surface corresponding to island x.

Input

  • n - the number of islands.
  • l,c - the dimensions of the matrix.
  • the next l lines, c numbers giving the lth row of the matrix.

Output

  • i - the number of the island with the largest area.
  • A - the area of the i‘th island.

Ex:

You have the following input:

2 4 4
0 0 0 1
0 0 1 1
0 0 0 2
2 2 2 2

For which you will get island no. 2 as the biggest island with the area of 5 squares.

Hints

The problem is quite easy, but here are some hints:

1. Use the flood-fill algorithm whenever you encounter a new island.
2. As opposed to the sample code, you should go through the area of the island and not on the ocean (0 tiles).