视觉惯性SLAM综述  

Overview of visual inertial SLAM

作　　者：孙永全 田红丽[2] Sun Yongquan;Tian Hongli(School of Artificial Intelligence,Hebei University of Technology,Tianjin 300401,China;Hebei Provincial Key Laboratory of Big Data Computing,Hebei University of Technology,Tianjin 300401,China)

机构地区：[1]河北工业大学人工智能与数据科学学院,天津300401 [2]河北工业大学河北省大数据计算重点实验室,天津300401

出　　处：《计算机应用研究》2019年第12期3530-3533,3552,共5页Application Research of Computers

基　　金：天津市科技计划资助项目(14ZCDGSF00124);天津市基础研究计划资助项目(17JCTPJC55400,17JCTPJC55600)

摘　　要：对视觉SLAM、惯性导航和视觉惯性SLAM进行了介绍,详细介绍了视觉SLAM的框架结构和细分部分,罗列了近年来出色的视觉SLAM模型。最后在视觉惯性里程计的(visual inertial odometry,VIO)基础上对现有的不同派别的开源视觉惯性SLAM进行深入分析与比较,探讨视觉惯性SLAM的发展趋势与动向。This paper researched visual SLAM,inertial navigation and visual inertia SLAM,and introduced some specific details about the frame structure and subdivision of visual SLAM. Besides,this paper also listed the excellent visual SLAM model in recent years. In the last part,it analyzed and compared the existing open source visual inertia SLAM of different factions based on visual inertial odometry( VIO),and discussed the development trend and trend of visual inertia SLAM.

关 键 词：视觉SLAM 视觉惯性SLAM 机器人视觉 惯性导航 视觉惯性里程计 

分 类 号：TP242.62[自动化与计算机技术—检测技术与自动化装置]