A Low-cost SLAM Fusion Algorithm for Robot Localization

• Author / Creator

  Ruifan Wu

• Autonomous navigation has been a popular research topic over the last two decades. The ability for a robot to solve the simultaneous localization and mapping (SLAM) problem is required to navigate unknown environments. One such example is a self-driving car: it needs to build maps of new environments and simultaneously localize itself. Many approaches for solving the SLAM problem have been proposed, but most of these require expensive LiDAR sen- sors. In this work, we investigate a strategy to solve the SLAM problem using an RGB-depth sensor Kinect v2, an open-source SLAM package called RTAB- Map, and an Extended Kalman Filter (EKF). Although SLAM algorithm we chose is computationally tractable and provides accurate results, the resulting pose data is available only at 1 Hz. To overcome the disadvantage of the low data frequency from RTAP-Map, we utilize an EKF filter to fuse in odometry estimates and obtain higher-rate (30 Hz) pose estimates. The system is imple- mented onboard a Jackal Unmanned Ground Vehicle (UGV), equipped with a Kinect v2 camera as the only external sensor. A motion capture system from Vicon is used to obtain a ground truth of the robot’s motions.Our results show good agreement between the pose estimates from our sys- tem and the ground truth. Our work demonstrates the ability of an inexpen- sive RGB-depth sensor such as the Kinect v2, combined with an open-source SLAM package and fused with high-rate odometry estimates through an EKF, to achieve good performance and accuracy results.

• Subjects / Keywords

  • SLAM
  • Robotic Localization

• Graduation date

  Fall 2019

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/r3-bk65-rm76

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.