ROS2 repo for running simulations and visualizing LiDAR data

This repository only contains the ros packages used for simulation and visualization purposes for the LiDAR based Object Detection and Tracking project. Kindly review the main repository to see the full project.

• Environment: Docker image
• ROS2 Humble
• Gazebo Fortress

NOTE: The ros2 pipeline developed here does not work on Xavier due to docker version issues and ROS2 Humble image on Jetson (dustynv) giving issues with ultralytics installation on Jetpack 5 (Ubuntu 2020.04).

NOTE Docker images with network-mode host work if the link-local setting of ethernet is properly configured. If you are using DHCP server (i.e. dhcp-kea or systemd-networkd) prefer using macvlan instead of host.

Commands

Running Simulations

# The sdf files are all stored in src/car_model/worlds, to add new simulation environment add it in src/car_model/worlds directory. The launch file is hardcoded to retrieve the sdf file from this path always i.e. world:=test_world.sdf will launch src/car_model/worlds/test_world.sdf file
# record tells to save the BEV video as avi file. This file can be later sent to Jetson for computing the bounding boxes.
# bag_file saves the recorded data in a bag file, if left empty it would not generate a bag file
ros2 launch car_model gazebo_model_launch world:=<path to sdf file> \
    record:=<true or false> \
    bag_file:=<name of bag file>

To control the car in the simulation environments use

# Using ros2
ros2 run teleop_twist_keyboard teleop_twist_keyboard

# using gazebo
ign topic -t "cmd_vel" -m ignition.msgs.Twist -p "linear: {x: 0.5}, angular: {z: 0.1}

Visualizing Live Sensor Data

# Rviz2 with all range, signal, NIR, and reflectivity images and 3D point cloud
ros2 launch ouster_ros sensor.launch.xml \
    sensor_hostname:=<sensor host name>

# Loading LiDAR properties via config file, see src/av_car/config/ouster_params.yaml
ros2 launch ouster_ros driver.launch.py \
    params_file:=<path to params yaml file>

# Wrapper for above function, loads Rviz2 with the src/av_car/config/ouster_params.yaml config file
ros2 launch av_car sensor_launch

# Live Sensor visualization with BEV image
ros2 launch av_car sensor_with_bev_launch.py \
    sensor_metadata:=metadata/ouster_metadata.json \
    sensor:=true

# Recorded sensor data visualization with BEV image
ros2 launch av_car sensor_with_bev_launch.py sensor:=false \
    sensor_metadata:=metadata/ouster_metadata.json \
    pcap_file:=recordings/pcap/test_pcap.pcap \
    pcap_metadata:=recordings/pcap/test_pcap_0.json

Storing Data

# ROSBAG
ros2 bag record -ao <filename>

ros2 bag record <topic_name> -o <filename>

# Recording Live sensor data in rosbag
ros2 launch ouster_ros record.launch.xml \
    sensor_hostname:=<sensor host name> \
    bag_file:=<optional bag file name> \
    metadata:=<json file name>

# Recording in both bag and pcap file simultaneously
ros2 launch av_car recording_launch \
    record_topic:=</lidar/points or /ouster/points> \ bag_file:=<bag_filename> \
    launch_type:=<sensor or gazebo> \ sensor_hostname:=<sensor_ip> \
    pcap_file:=<pcap_file_name>

# PCAP
ouster-cli source <sensor_hostname> save <filename>.pcap

Replaying Data

# BAG
ros2 bag replay <bag_file_dir>

# PCAP
ros2 launch ouster_ros replay_pcap.launch.xml \
    pcap_file:=<path to ouster pcap file> \
    metadata:=<json file name>