Effective method of Initial Seed Selection Used for Color Image Segmentation

We propose a method of initial seed selection used for color image segmentation. In this paper Initial seed selection is a starting stage of color image segmentation. This method is based on HSV color model. Here we use two details; one is non-edge pixels and second is smoothness feature of the pixels. The non-edge pixels and smoothness at pixel’s neighbor are used as criteria to determine initial seeds in the color image. No-edge information provide those pixels are not on the edge and smoothness provides pixels have high similarity to its neighbors. This method is applied to many color images and experimental results show the effectiveness of the method.