Remote sensing image segmentation using RPCCL and labeling watershed transform
Xinhui Li, Renxi Chen · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2009
Image segmentation is the technology that separating the image into several characteristic areas and is very important to image analysis. In this paper, we propose a new segmentation method based on rival penalized controlled competitive learning (RPCCL) and watershed transform. We apply watershed transform and RPCCL clustering algorithm separately on input image, and then combine the two results of them. Compared to traditional watershed segmentation, our method can avoid over-segmentation and obtain better results.