OKVIS: Open Keyframe-based Visual-Inertial SLAM (ROS Version)
Switch branches/tags
Clone or download
Latest commit bdc6db6 Jul 22, 2016

README.md

README {#mainpage}

Welcome to OKVIS: Open Keyframe-based Visual-Inertial SLAM.

This is the Author's implementation of the [1] and [3] with more results in [2].

[1] Stefan Leutenegger, Simon Lynen, Michael Bosse, Roland Siegwart and Paul Timothy Furgale. Keyframe-based visual–inertial odometry using nonlinear optimization. The International Journal of Robotics Research, 2015.

[2] Stefan Leutenegger. Unmanned Solar Airplanes: Design and Algorithms for Efficient and Robust Autonomous Operation. Doctoral dissertation, 2014.

[3] Stefan Leutenegger, Paul Timothy Furgale, Vincent Rabaud, Margarita Chli, Kurt Konolige, Roland Siegwart. Keyframe-Based Visual-Inertial SLAM using Nonlinear Optimization. In Proceedings of Robotics: Science and Systems, 2013.

Note that the codebase that you are provided here is free of charge and without any warranty. This is bleeding edge research software.

Also note that the quaternion standard has been adapted to match Eigen/ROS, thus some related mathematical description in [1,2,3] will not match the implementation here.

If you publish work that relates to this software, please cite at least [1].

How do I get set up?

This is a catkin package that wraps the pure CMake project.

You will need to install the following dependencies,

  • ROS (currently supported: hydro, indigo and jade). Read the instructions in http://wiki.ros.org/indigo/Installation/Ubuntu. You will need the additional package pcl-ros as (assuming indigo)

      sudo apt-get install ros-indigo-pcl-ros
    
  • google-glog + gflags,

      sudo apt-get install libgoogle-glog-dev
    
  • The following should get installed through ROS anyway:

      sudo apt-get install libatlas-base-dev libeigen3-dev libsuitesparse-dev 
      sudo apt-get install libopencv-dev libboost-dev libboost-filesystem-dev
    
  • Optional: use the the package with the Skybotix VI sensor. Note that this requires a system install, not just as ROS package. Also note that Skybotix OSX support is experimental (checkout the feature/osx branch).

      git clone https://github.com/ethz-asl/libvisensor.git
      cd libvisensor
      ./install_libvisensor.sh
    

then download and expand the archive into your catkin workspace:

wget https://www.doc.ic.ac.uk/~sleutene/software/okvis_ros-1.1.3.zip
unzip okvis_ros-1.1.3.zip && rm okvis_ros-1.1.3.zip

Or, clone the repository from github into your catkin workspace:

git clone --recursive git@github.com:ethz-asl/okvis_ros.git

or

git clone --recursive https://github.com/ethz-asl/okvis_ros.git

Building the project

From the catkin workspace root, type

catkin_make

You will find a demo application in okvis_apps. It can process datasets in the ASL/ETH format.

In order to run a minimal working example, follow the steps below:

  1. Download a dataset of your choice from http://projects.asl.ethz.ch/datasets/doku.php?id=kmavvisualinertialdatasets. Assuming you downloaded MH_01_easy/. You will find a corresponding calibration / estimator configuration in the okvis/config folder.

  2. Run the app as

     ./okvis_apps path/to/okvis_ros/okvis/config/config_fpga_p2_euroc.yaml path/to/MH_01_easy/
    

You can also run a dataset processing ros node that will publish topics that can be visualized with rviz

rosrun okvis_ros okvis_node_synchronous path/to/okvis_ros/okvis/config/config_fpga_p2_euroc.yaml path/to/MH_01_easy/

Use the rviz.rviz configuration in the okvis_ros/config/ directory to get the pose / landmark display.

If you want to run the live application connecting to a sensor, use the okvis_node application (modify the launch file launch/okvis_node.launch).

Outputs and frames

In terms of coordinate frames and notation,

  • W denotes the OKVIS World frame (z up),
  • C_i denotes the i-th camera frame,
  • S denotes the IMU sensor frame,
  • B denotes a (user-specified) body frame.

The output of the okvis library is the pose T_WS as a position r_WS and quaternion q_WS, followed by the velocity in World frame v_W and gyro biases (b_g) as well as accelerometer biases (b_a). See the example application to get an idea on how to use the estimator and its outputs (callbacks returning states).

The okvis_node ROS application will publish a configurable state -- see just below.

Configuration files

The okvis/config folder contains example configuration files. Please read the documentation of the individual parameters in the yaml file carefully. You have various options to trade-off accuracy and computational expense as well as to enable online calibration. You also have various options concerning the things that will get published -- in particular weather or not landmarks should be published (may be important to turn off for on-bard operation). Moreover, you can specify how the body frame is specified (T_BS) or define a custom World frame. In other words, the final pose published will be T_Wc_B = T_Wc_W * T_WS * T_BS^(-1) . You have the option to express the velocity as well as the rotation rates in either B, S, or Wc.

HEALTH WARNING: calibration

If you would like to run the software/library on your own hardware setup, be aware that good results (or results at all) may only be obtained with appropriate calibration of the

  • camera intrinsics,
  • camera extrinsics (poses relative to the IMU),
  • knowledge about the IMU noise parameters,
  • and ACCURATE TIME SYNCHRONISATION OF ALL SENSORS.

To perform a calibration yourself, we recommend the following:

Contribution guidelines

Support

The developpers will be happy to assist you or to consider bug reports / feature requests. But questions that can be answered reading this document will be ignored. Please contact s.leutenegger@imperial.ac.uk.