About

• Object residual constrained Visual-Inertial Odometry (OrcVIO) is a visual-inertial odometry pipeline, which is tightly coupled with tracking and optimization over structured object models. It provides accurate trajectory estimation and large-scale object-level mapping from online Stereo+IMU data.

• OrcVIO-Lite only uses bounding boxs and no keypoints. The object mapping module and VIO module are implemented in separate ROS nodelets and are decoupled.

• Project website

Citation

@inproceedings{shan2020orcvio,
  title={OrcVIO: Object residual constrained Visual-Inertial Odometry},
  author={Shan, Mo and Feng, Qiaojun and Atanasov, Nikolay},
  booktitle={2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages={5104--5111},
  year={2020},
  organization={IEEE}
}   

1. Prerequisites

This repository was tested on Ubuntu 18.04 with ROS Melodic.

The core algorithm depends on Eigen, Boost, Suitesparse, Ceres, OpenCV, Sophus, GTest

2. Installation

ROS version

• Environment is Ubuntu 18.04 with ROS Melodic
• The ROS version also depends on catkin simple, please put it in the ros_wrapper/src folder

$ git clone --recursive https://github.com/shanmo/OrcVIO-Stereo-Mapping.git
$ cd OrcVIO-Stereo-Mapping/ros_wrapper
$ catkin_make
$ source ./devel/setup.bash

3. Demo

• Download ERL indoor hand-held dataset (chairs), which was collected with Realsense D455 in Existential Robotics Lab, University of California San Diego

• Please refer to wiki regarding how to setup D455

• Run roslaunch orcvio orcvio_d455.launch generates the result below

• Download ERL indoor robotic car dataset, change the rosbag path, and run roslaunch orcvio orcvio_d455.launch to get the demos like below

License

MIT License
Copyright (c) 2021 ERL at UCSD

Reference

• LARVIO
• MSCKF
• OpenVINS
• OpenVINS eval