Global Visual–Inertial Localization for Autonomous Vehicles with Pre-Built Map
Open Access
Abstract
Accurate, robust and drift-free global pose estimation is a fundamental problem for autonomous vehicles. In this work, we propose a global drift-free map-based localization method for estimating the global poses of autonomous vehicles that integrates visual–inertial odometry and global localization with respect to a pre-built map. In contrast to previous work on visual–inertial localization, the global pre-built map provides global information to eliminate drift and assists in obtaining the global pose. Additionally, in order to ensure the local odometry frame and the global map frame can be aligned accurately, we augment the transformation between these two frames into the state vector and use a global pose-graph optimization for online estimation. Extensive evaluations on public datasets and real-world experiments demonstrate the effectiveness of the proposed method. The proposed method can provide accurate global pose-estimation results in different scenarios. The experimental results are compared against the mainstream map-based localization method, revealing that the proposed approach is more accurate and consistent than other methods.Keywords
Funding Information
- Beijing Natural Science Foundation (JQ20013, 61833009, 62273195, U19B2029)
- National Natural Science Foundation of China (JQ20013, 61833009, 62273195, U19B2029)
This publication has 32 references indexed in Scilit:
- On-Manifold Preintegration for Real-Time Visual--Inertial OdometryIEEE Transactions on Robotics, 2016
- The EuRoC micro aerial vehicle datasetsThe International Journal of Robotics Research, 2016
- Robust visual inertial odometry using a direct EKF-based approach2015
- Real-time visual-inertial localization for aerial and ground robots2015
- Keyframe-based visual–inertial odometry using nonlinear optimizationThe International Journal of Robotics Research, 2014
- Bags of Binary Words for Fast Place Recognition in Image SequencesIEEE Transactions on Robotics, 2012
- Improving the accuracy of EKF-based visual-inertial odometry2012
- Real-time metric state estimation for modular vision-inertial systems2011
- Visual-inertial navigation, mapping and localization: A scalable real-time causal approachThe International Journal of Robotics Research, 2011
- A Multi-State Constraint Kalman Filter for Vision-aided Inertial NavigationProceedings 2001 ICRA. IEEE International Conference on Robotics and Automation (cat. No.01ch37164), 2007