Perceptual Segmentation of Natural Color Images by Fuzzy-Based Split-and-Merge Algorithm and It's Evaluation

Abstract

The purpose of this paper is to develop the perceptual segmentation algorithm of natural color images, and to quantitatively evaluate the algorithm using ground truth images that are human-segmented results. We propose the fuzzy-based split-and-merge algorithm as perceptual segmentation method that integrates color features and texture features by using fuzzy inference. We perform the comparative experiments of the proposed method in comparison with the other methods to demonstrate the effectiveness of the proposed method, in which the Normalized Probabilistic Rand Index and the Dongen Index are used as the measures of evaluation with ground truth images.