In this paper the implementation of a parallel watershed algorithm is described. The algorithm is implemented on a multiple instruction multiple data (MIMD) ring-architecture using a single program multiple data (SPMD) approach using an asynchronous message passing interface and simulated shared memory via the Linda tuple space. The watershed transform is generally considered to be inherently sequential. This paper shows that it is possible to exploit parallelism by splitting the computation of the watersheds of an image into three stages that can be executed in parallel.
|Title of host publication||EPRINTS-BOOK-TITLE|
|Publisher||University of Groningen, Johann Bernoulli Institute for Mathematics and Computer Science|
|Number of pages||9|
|Publication status||Published - 1995|
- image segmentation
- parallel implementation
- watershed algorithm