WATERSHEDS IN DIGITAL SPACES - AN EFFICIENT ALGORITHM BASED ON IMMERSION SIMULATIONS

In this paper, a fast and flexible algorithm for computing watersheds in digital grayscale images is introduced. A review of watersheds and related notion is first presented, and the major methods to determine watersheds are discussed. The present algorithm is based on an immersion process analogy, in which the flooding of the water in the picture is efficiently simulated using a queue of pixels. It is described in detail and provided in a pseudo C language. We prove the accuracy of this algorithm is superior to that of the existing implementations. Furthermore, it is shown that its adaptation to any kind of digital grid and its generalization to n-dimensional images and even to graphs are straightforward. In addition, its strongest point is that it is faster than any other watershed algorithm. Applications of this algorithm with regard to picture segmentation are presented for MR imagery and for digital elevation models. An example of 3-D watershed is also provided. Lastly, some ideas are given on how to solve complex segmentation tasks using watersheds on graphs.