A Cooperative Relaxation-Based Method for Range Image Segmentation

Abstract

In this work, we propose a combination of two methods for the segmentation of range images (Depth Image), based on the relaxation of the detection results of both edges and regions. The latter are obtained from two different representations of the data. The principle of the introduced method consists in matching the two detection results as a mutual self-regularization, and whose objective is to produce an optimal final segmentation with respect to a given criterion of optimality, expressed by a well appropriate objective function. The experimental results, using the ABW database, show the potential of the proposed method for efficient segmentation of range images.