Relational graph location network for multi-view image localization

作     者：YANG Yukun LIU Xiangdong YANG Yukun;LIU Xiangdong

作者机构：School of AutomationBeijing Institute of TechnologyBeijing 100081China 

出 版 物：《Journal of Systems Engineering and Electronics》 (系统工程与电子技术（英文版）)

年 卷 期：2023年第34卷第2期

页      面：460-468页

核心收录：

学科分类：08[工学] 080203[工学-机械设计及理论] 0802[工学-机械工程] 

主　　题：multi-view image localization graph construction heterogeneous graph graph neural network 

摘      要：In multi-view image localization task,the features of the images captured from different views should be fused *** paper considers the classification-based image localization *** propose the relational graph location network(RGLN)to perform this *** this network,we propose a heterogeneous graph construction approach for graph classification tasks,which aims to describe the location in a more appropriate way,thereby improving the expression ability of the location representation *** show that the expression ability of the proposed graph construction approach outperforms the compared methods by a large *** addition,the proposed localization method outperforms the compared localization methods by around 1.7%in terms of meter-level accuracy.