Application of Combining Watershed and Fast Clustering Method in Image Segmentation

Watershed algorithm is an image segmentation method based on mathematical morphology, it can realize parallel regional partition and get complete segmentation regionals. Watershed algorithm has sensitive effect on weak edge, but this method easily products the over segmentation which results in the edge line buried in disorderly watershed lines. To avoid over segmentation produced by watershed segment algorithm, a new image segment algorithm W-SPK (combining watershed and K-means clustering based on simulated annealing particle swarm optimization) is proposed in this paper. W-SPK effectively overcomes the lack of watershed algorithm, and it makes full use of the ability of global optimization of PSO (particle swarm optimization) and the advantage of jumping out local minimum of SA (simulated annealing). The experimental results indicate that W-SPK can not only basically eliminates the over-segmentation problem, but also has the character of fast and efficient clustering. W-SPK is a rapid and accurate method to realize significant image segmentation.