Spatial structure analysis for autonomous robotic vision systems

Abstract

Analysis of spatial structures in robotic environments, especially structures such as planar surfaces, has become a fundamental component in diverse robot vision systems since the introduction of low-cost RGB-D cameras that have been widely mounted on various indoor robots. These cameras are capable of providing high quality 3D reconstruction in real time. In order to estimate multiple planar structures without prior knowledge, this paper utilizes Jensen-Shannon Divergence (JSD), which is a similarity measurement method, to represent pairwise relationship between data. This conceptual representation encompasses the pairwise geometrical relations between data as well as the information about whether pairwise relationships exist in a model's inlier data set or not. Tests on datasets comprised of noisy inliers and a large percentage of outliers demonstrate that the proposed solution can efficiently estimate multiple models without prior information. Superior performance in terms of synthetic experiments and pragmatic tests with robot vision system also demonstrate the validity of the proposed approach.