Robust Monocular Visual-Inertial SLAM Using Nonlinear Optimization

  • Jingyun Duo
  • , Lei Ji
  • , Long Zhao*
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, a robust monocular visual-inertial SLAM based on nonlinear optimization is proposed. In our method, visual feature points are assigned different information matrices according to the image pyramid layers at which the features are extracted. IMU pre-integration strategy is adopted to avoid repeated IMU integration caused by initial states change in optimization. Meanwhile, we adopted the strategies of sliding window and marginalization in order to yield higher precision of states estimation and restrict the computational complexity. Experiments are designed to compare our algorithm with MSCKF and VINS on EuRoC dataset, and the results show that our method can effectively estimate the motion and sparse map.

Original languageEnglish
Title of host publicationProceedings of 2020 Chinese Intelligent Systems Conference - Volume I
EditorsYingmin Jia, Weicun Zhang, Yongling Fu
PublisherSpringer Science and Business Media Deutschland GmbH
Pages560-568
Number of pages9
ISBN (Print)9789811584497
DOIs
StatePublished - 2021
EventChinese Intelligent Systems Conference, CISC 2020 - Shenzhen, China
Duration: 24 Oct 202025 Oct 2020

Publication series

NameLecture Notes in Electrical Engineering
Volume705 LNEE
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

Conference

ConferenceChinese Intelligent Systems Conference, CISC 2020
Country/TerritoryChina
CityShenzhen
Period24/10/2025/10/20

Keywords

  • Inertial
  • Monocular visual
  • Nonlinear optimization
  • SLAM
  • State estimation

Fingerprint

Dive into the research topics of 'Robust Monocular Visual-Inertial SLAM Using Nonlinear Optimization'. Together they form a unique fingerprint.

Cite this