A Novel Approach for Image Segmentation Enhancement through Multi-Region Clustering Technique

Image segmentation has been, and still is, a relevant research area in Computer Vision, and hundreds of segmentation algorithms have been implemented in the last 30 years. However, it is well known that elemental segmentation techniques based on boundary or region information often fail to produce accurate segmentation results and efficiency. This paper focused on three segmentation techniques i.e Stochastic Random Walker Method, Reflection Symmetry, Spatially coherent Fuzzy Clustering Method, which integrate edge and region information to fused such information. It is also useful for feature extraction, image measurements and image display but reduces speed, and efficiency. To improve the speed of processing drastically, improved image segmentation by exploiting image symmetry, gives a reliability, preserve efficiency, so we propose Multi-region clustering algorithm (MRC). Segmentation using multi-region clustering can perform well in intersection areas in comparison with manual ones. The propose method can achieve satisfactory result.