基于二次曲面拓扑关系的工件位姿估算方法  

A Method of Workpiece Pose Estimation Based on Quadric Topological Relation

作　　者：徐进 李德平[1,2] 柳宁 Xu Jin;Li Deping;Liu Ning(College of Information Science and Technology,Jinan University;Robotics Research Institute of Jinan University)

机构地区：[1]暨南大学信息科学技术学院 [2]暨南大学机器人智能技术研究院

出　　处：《自动化与信息工程》2019年第6期26-32,共7页Automation & Information Engineering

基　　金：国家自然科学基金（61775172）;广东省自然科学基金（2018030310482）

摘　　要：针对Bin-Picking系统中工件6自由度位姿估算,基于全局和局部特征的点云匹配算法,对工业零件存在的一定局限性问题,提出一种基于二次曲面拓扑关系的工件位姿估算方法。该方法考虑到工业零件表面的曲面特征,利用图结构描述曲面间拓扑关系;并通过子图同构匹配完成目标对象的识别;最后利用曲面特征参数进行快速的位姿估算。实验证明,该方法比传统位姿估算方法更快速、准确,具有良好的鲁棒性和可拓展性,能满足工业实时性要求。For Bin-Picking system of workpiece 6-Dof position matching problem, point cloud based on global and local feature matching algorithm for industrial components has certain limitation, then put forward a kind of topological relations based on quadric surface scattered workpiece position estimation techniques, the proposed algorithm considering the surface characteristics of industrial components surface, graph is used to describe topological relations between surface and through the subgraph isomorphism matching complete target recognition, the use of surface characteristic parameters for pose estimation. Experimental results show that the method we proposed is faster and more accurate than traditional pose matching algorithm, and has good robustness and expansibility.

关 键 词：Bin-Picking 点云匹配 二次曲面 子图同构 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TG95[自动化与计算机技术—计算机科学与技术]