A Multi-Scale Image Segmentation Algorithm Based on the Cloud Model

Abstract

A new method of image segmentation based on cloud model theory is proposed in this paper. A major contribution of this work is to add uncertainty of image to the segmentation algorithm. Segmentation is realized in three stages. First, we use cloud model theory to transform the image’s qualitative model to its quantitative model (concept tree). Second, we use climbing policy to get different level concepts which represent different level objects. At last, determine which concept each pixel belongs to. Such process will generate a scale-space hierarchical tree that induces segmentation without a priori knowledge. Experimental results based on natural images with respect to the concept tree and segmentation proved this multi-scale image segmentation algorithm can get different levels objects very well and good at resolving the edges of different objects which have uncertainty to objects.