Image Segmentation Based on Color Quantization Using Self-Organizing Map

Abstract

This paper presents the color quantization of an image using a self-organizing map (SOM) and image segmentation. We try to implement perceptual segmentation of color images using only color quantization without using texture features. We propose an improved SOM that is a modified version of the Ong's method. We also propose to introduce the pre-processing including the correction of achromatic regions based on fracral dimension and the edge-preserving smoothing by using anisotropic diffusion, that is suitable for reducing small regions and realizing better segmentation than the conventional methods.