MRGBD-VIO

A Multi-Feature tightly-coupled RGB-D visual-inertial SLAM system. The proposed system is the first tightly coupled optimization-based RGB-D-inertial system based on multi-features. This system is runs on Linux and ROS. Based on the open source SLAM framework VINS-Mono.

1. Prerequisites

1.1 Ubuntu and ROS Ubuntu 16.04. ROS Kinetic, ROS Installation additional ROS pacakge

    sudo apt install ros-Kinetic-desktop-full

1.2 Opencv3

If you install ROS Kinetic, please update opencv3 with

    sudo apt-get install ros-kinetic-opencv3

1.3 Ceres Solver Follow Ceres Installation, remember to make install.

1.4 Sophus

    git clone http://github.com/strasdat/Sophus.git

2. Build MRGBD-VIO on ROS

Clone the repository and catkin_make:

    cd ~/catkin_ws/src
    git clone https://github.com/Grandzxw/MRGBD-VIO.git
    cd ../
    catkin_make
    source ~/catkin_ws/devel/setup.bash

3.Run on OpenLORIS dataset

  roslaunch vins_estimator realsense_color.launch
  roslaunch vins_estimator vins_rviz.launch
  rosbag play bagname.bag

4. OpenLORIS dataset

• OpenLORIS

5. Citation

@article{zhao2020multi,
  title={Multi-Feature Nonlinear Optimization Motion Estimation Based on RGB-D and Inertial Fusion},
  author={Zhao, Xiongwei and Miao, Cunxiao and Zhang, He},
  journal={Sensors},
  volume={20},
  number={17},
  pages={4666},
  year={2020},
  publisher={MDPI}
}

6. Licence

The source code is released under GPLv3 license.