Fejemis ROS

Back to fejemis

ROS2 Iron

A new ROS2 installation is in progress.

The following packages are needed.

• ros-iron-gazebo
• ros-iron-xacru
• ros-iron-ros2-control
• ros-iron-ros2-controllers
• ros-iron-gazebo-ros2-control
• ros-iron-ros2-slam-toolbox
• ros-iron-twist-mux

The base repository is pt.: https://github.com/Lohmann27/fejemis_sim.git

ROS install (depricated)

The version of ROS used on Fejemis is ROS noetic.

An installing guide can be found here: ROS Noetic installation instructions

The version installed is the Desktop-Full Install.

Catkin tools

Catkin build is used in the development and therefore catkin_tools need to be installed.

On Ubuntu apt-get can be used via this guide: Installing on Ubuntu with apt-get

Intel RealSense

The installation also uses Intel RealSense and can be installed via this guide. Instructions

Install both realsense2_camera and realsense2_description

Calibration of RealSense

For calibration on linux install this: Installation Instructions for librealsense2-dkms and librealsense2-utils

IMU and robot-localization

Also for the IMU and integration with RTAB-Map, we use robot_localization

Use this

sudo apt-get install ros-noetic-robot-localization

and this

sudo apt-get install ros-noetic-imu-tools

Rviz Add-on (visualization)

For the Rviz visualization when using Intel RealSense install this.

sudo apt-get install ros-noetic-octomap-rviz-plugins

RTAB-Map

Standard install

sudo apt install ros-noetic-rtabmap-ros

Info about dual camera setup

Install Git version of RTAB-map

If you need both camera for SLAM the normal install will not work. Install from Github like this.

sudo apt remove ros-$ROS_DISTRO-rtabmap-ros

cd ~/catkin_ws/src

git clone -b $ROS_DISTRO-devel https://github.com/introlab/rtabmap_ros.git

cd ~/catkin_ws

catkin build -DRTABMAP_SYNC_MULTI_RGBD=ON

Get the serial numbers for the cameras

rs-enumerate-devices | grep 'Serial Number'

More info

[1](https://answers.ros.org/question/320708/rtabmap-ros-with-two-realsense-d435-cameras/)

You need to of the same cameras, as the size of the pictures should be the same.

Another way of implementing it

[2](https://answers.ros.org/question/269459/rtabmap-with-two-stereo-cameras/)

(here we just update the obstacle layer of the cost maps with camera 2)