Dedicated package to demonstrate gazebo camera timestamp issue https://bitbucket.org/osrf/gazebo/issues/1748/significant-sensor-data-timestamp-accuracy with a minimal example
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
config
launch
rviz
urdf
CMakeLists.txt
README.md
package.xml

README.md

gazebo_camera_timestamp_issue

Dedicated package to demonstrate gazebo camera timestamp issue https://bitbucket.org/osrf/gazebo/issues/1748/significant-sensor-data-timestamp-accuracy with a minimal example

Installation

The package can be added to a ROS workspace and works without any further dependencies, provided the ROS packagesgazebo_ros_control, ros_controllers and gazebo_plugins are available. Easiest is to get them via apt-get:

sudo apt-get install ros-$ROS_DISTRO-gazebo-ros-control ros-$ROS_DISTRO-ros-controllers ros-$ROS_DISTRO-gazebo-plugins 

Test Setup

The test setup consists of a single URDF model that is equipped with a depth sensor mounted on a continuously spinning joint. Along the x axis, a pole is mounted in exactly 1m distance (distance between camera and closest point on pole). To the left, along the y axis, a box is mounted. The test can be started by running

roslaunch gazebo_camera_timestamp_issue run_test.launch

A rviz window shows sensor data and robot model. The expected/intended result is that sensor data aligns with the robot model. This is what happens when using Gazebo2 as shown in this video:

Alt text

With later Gazebo versions, the intended behavior is not observed anymore and there is a significant offset, with the most likely cause appearing to be an offset between camera and joint timestamps:

Alt text