ROS Package

We provide ROS package examples to help you run OpenVSLAM on ROS framework.

Note

Please build OpenVSLAM with OpenCV 3.3.1 or later if you plan on using ROS package. OpenCV 4.x is not supported yet.

Installation

Requirements

ROS : Please use the version kinetic or later.
OpenVSLAM : Please build it with OpenCV 3.x.
image_transport : Required by this ROS package examples.
cv_bridge : Please build it with the same version of OpenCV used in OpenVSLAM. (We recommend building it from source.)

Prerequisites

Tested for Ubuntu 16.04.

Please install the following dependencies.

ROS : Please follow Installation of ROS.
OpenVSLAM : Please follow Installation of OpenVSLAM.

Note

Please build OpenVSLAM with PangolinViewer or SocketViewer if you plan on using it for the examples.

Install the dependencies via apt.

apt update -y
apt install ros-${ROS_DISTRO}-image-transport

Download the source of cv_bridge.

cd /path/to/openvslam/ros
git clone --branch ${ROS_DISTRO} --depth 1 https://github.com/ros-perception/vision_opencv.git
cp -r vision_opencv/cv_bridge src/
rm -rf vision_opencv

Note

We recommend building cv_bridge from the source even if it has been installed via apt.

Build Instructions

When building with support for PangolinViewer, please specify the following cmake options: -DUSE_PANGOLIN_VIEWER=ON and -DUSE_SOCKET_PUBLISHER=OFF as described in build of OpenVSLAM.

cd /path/to/openvslam/ros
catkin_make \
    -DBUILD_WITH_MARCH_NATIVE=ON \
    -DUSE_PANGOLIN_VIEWER=ON \
    -DUSE_SOCKET_PUBLISHER=OFF \
    -DUSE_STACK_TRACE_LOGGER=ON \
    -DBOW_FRAMEWORK=DBoW2

Alternatively, when building with support for SocketViewer, please specify the following cmake options: -DUSE_PANGOLIN_VIEWER=OFF and -DUSE_SOCKET_PUBLISHER=ON as described in build of OpenVSLAM.

cd /path/to/openvslam/ros
catkin_make \
    -DBUILD_WITH_MARCH_NATIVE=ON \
    -DUSE_PANGOLIN_VIEWER=OFF \
    -DUSE_SOCKET_PUBLISHER=ON \
    -DUSE_STACK_TRACE_LOGGER=ON \
    -DBOW_FRAMEWORK=DBoW2

Examples

Run the core program required for ROS-based system in advance.

roscore

Note

Please leave the roscore run.

Publisher

Publishers continually broadcast images as a ROS topic. Please execute one of the following command snippets in the new terminal.

Publish a Video File

For using video files (e.g. .mp4) for visual SLAM or localization.

source /path/to/openvslam/ros/devel/setup.bash
rosrun publisher video -m /path/to/video.mp4

Republish the ROS topic to /camera/image_raw.

rosrun image_transport republish \
    raw in:=/video/image_raw raw out:=/camera/image_raw

Publish a Image Sequence

For using image sequences for visual SLAM or localization.

source /path/to/openvslam/ros/devel/setup.bash
rosrun publisher image -i /path/to/images/

Republish the ROS topic to /camera/image_raw.

rosrun image_transport republish \
    raw in:=/video/image_raw raw out:=/camera/image_raw

Publish Images of a USB Camera

For using a standard USB camera for visual SLAM or localization.

apt install ros-${ROS_DISTRO}-usb-cam

rosparam set usb_cam/pixel_format yuyv
rosrun usb_cam usb_cam_node

Republish the ROS topic to /camera/image_raw.

rosrun image_transport republish \
    raw in:=/usb_cam/image_raw raw out:=/camera/image_raw

Subscriber

Subscribers continually receive images. Please execute one of the following command snippets in the new terminal.

Note

Option arguments are the same as the examples of OpenVSLAM.

Tracking and Mapping

We provide an example snippet for visual SLAM. The source code is placed at ./openvslam/ros/src/openvslam/src/run_slam.cc.

source /path/to/openvslam/ros/devel/setup.bash
rosrun openvslam run_slam \
    -v /path/to/orb_vocab.dbow2 \
    -c /path/to/config.yaml

Localization

We provide an example snippet for localization based on a prebuilt map. The source code is placed at ./ros/src/openvslam/src/run_localization.cc.

source /path/to/openvslam/ros/devel/setup.bash
rosrun openvslam run_localization \
    -v /path/to/orb_vocab.dbow2 \
    -c /path/to/config.yaml \
    --map-db /path/to/map.msg