RealSense-ROS2-Raspberry-Pi-4

A step-by-step instruction on how to set up RealSense ROS2 node on Raspberry Pi 4

Prerequisites

• Ubuntu Server 22.04 LTS: Use rpi-imager to install Ubuntu Server 22.04 LTS 64bit version to the Pi4.

• ROS2 Humble: Follow the official instructions to install ROS2 on the Pi4.

• Dependencies: Install necessary packages based on the instructions given by the official documentation

    sudo apt-get install \
    	libssl-dev libusb-1.0-0-dev libudev-dev pkg-config libgtk-3-dev \
    	git wget cmake build-essential \
    	libglfw3-dev libgl1-mesa-dev libglu1-mesa-dev at
  

• RealSense Camera Firmware version: It is recommended to use the realsense-viewer on a x86 computer to upgrade or downgrade the D435i camera's firmware to version 5.15.0.2 as it is better compatible with the SDK version 2.54.1.

• Disable Pi4's embedded camera devices:

    sudo modprobe -r bcm2835_codec
    sudo modprobe -r bcm2835_isp
  

RealSense SDK

In the home folder, download the RealSense SDK:

cd ~ && git clone -b v2.54.1 https://github.com/IntelRealSense/librealsense.git

Inside the RealSense SDK repository directory, unplug the camera and set up udev rules:

cd ~/librealsense && ./scripts/setup_udev_rules.sh

Create the build folder:

mkdir build && cd build

Configure the SDK source code:

cmake \
-DBUILD_GRAPHICAL_EXAMPLES=false \
-DBUILD_PYTHON_BINDINGS=true \
-DPYTHON_EXECUTABLE=/usr/bin/python3 \
-DFORCE_RSUSB_BACKEND=true \
-DCMAKE_BUILD_TYPE=Release ../

Compile and install:

make -j4 all && sudo make install

RealSense ROS2

In the home folder, create a ros2 workspace:

cd ~ && mkdir -p ros2_ws/src && cd ros2_ws/src

Download the RealSense ROS repository:

git clone https://github.com/IntelRealSense/realsense-ros.git && cd ..

Build the ROS2 package:

colcon build

Run the node

Source corresponding environments:

source ~/ros2_ws/install/setup.bash

Run the node with an acceptable resolution (acceptable for the Pi):

ros2 launch realsense2_camera rs_launch.py \
depth_module.profile:=424x240x6 \
rgb_camera.profile:=424x240x6 \
pointcloud.enable:=true

Visualise the point cloud

On the host computer with ROS2 installed, open RVIZ2:

ros2 run rviz2 rviz2

On the left panel, set the Fixed Frame to camera_link and add the topic /camera/camera/depth/color/points.

Tips

One can try with higher resolution but it might cause the node to freeze.

There will be a lot of warnings like (messenger-libusb.cpp:42) control_transfer returned error, index: 768, error: Resource temporarily unavailable, number: 11 even when the node is working.

The CMake flag 'FORCE_RSUSB_BACKEND' must be enabled or the ROS node will report error.