Skip to content

README.md

Latest commit

 

History

History
81 lines (53 loc) · 3.39 KB

README.md

File metadata and controls

81 lines (53 loc) · 3.39 KB

Ouster OS1 Sensor Description and Visualization

A minimal ROS package to visualize an Ouster OS1-64 sensor model in RViz against an existing, namespaced TF tree.

Overview

This package provides a simple way to display the 3D model of an Ouster OS1-64 sensor in RViz. It is specifically designed for situations where another ROS node or a rosbag is already publishing the sensor's coordinate frames (TFs), especially when those frames are published with a ROS namespace (e.g., /os1/os_sensor).

It correctly handles the common problem of linking a URDF model (which uses a tf_prefix with underscores, like os1_lidar) to an existing TF tree that uses namespaces (like os1/os_lidar).

Prerequisites

  • ROS Noetic (or a similar version with catkin).
  • A running roscore.
  • An active data source (e.g., a rosbag play command or a live driver) that is publishing the sensor's TF frames.

Installation

  1. Clone this repository into your catkin workspace's src directory:

    cd ~/your_catkin_ws/src/
git clone <URL_to_your_repository>

  2. Build the workspace:

    cd ~/your_catkin_ws/
catkin_make

  3. Source the workspace to make the package available in your environment:

    source devel/setup.bash

Usage

This package is designed to run alongside your existing data source. You will typically use two terminals.

Terminal 1: Your Data Source

In your first terminal, start the node or rosbag that publishes your sensor's data and TF frames.

Example using rosbag play:

# This bag must publish TF frames like /os1/os_sensor, /os1/os_lidar, etc.
rosbag play your_ouster_data.bag --clock

Terminal 2: Launch the Visualization

In your second terminal, run the launch file from this package.

roslaunch ouster_description visualize.launch tf_prefix:=os1

The tf_prefix Argument

The tf_prefix argument is crucial. It must match the namespace used in your TF data.

  • If your frames are os1/os_sensor, os1/os_lidar -> use tf_prefix:=os1
  • If your frames are my_robot/sensor_link, my_robot/lidar_link -> use tf_prefix:=my_robot

You should now see an RViz window with the Ouster sensor model correctly positioned relative to your other TF frames.

How It Works

This package uses a few key components to solve the visualization problem:

  1. urdf/my_ouster_model.urdf.xacro: This is the top-level robot description file. It defines a dummy world link and instantiates the Ouster model from OS1-64.urdf.xacro, attaching it to this world link.

  2. launch/visualize.launch: This is the main entry point. It does three things:

    • Loads the URDF from my_ouster_model.urdf.xacro to the /robot_description parameter.
    • Starts the robot_state_publisher to publish the internal transforms of the sensor model (e.g., the transform from os1 to os1_imu).
    • Starts a static_transform_publisher to create a "bridge".

The TF Bridge

The key to this package is the static_transform_publisher node launched in visualize.launch. It publishes a single, static transform that connects the namespaced frame from your data source (e.g., os1/os_sensor) to the world link defined in our URDF.

This bridge is what solves the "No transform from [world] to [base_footprint]" error in RViz, snapping the sensor model into the correct place within your robot's main TF tree.