ROS Lidar Translator Tutorial

This article was written for version 2.0.9 of Core. Click here for the most recent version of the help center.

This tutorial will show you how to build a basic ROS node that can take in custom PolySync messages in ROS, and republish them as native ROS types. It also demonstrates how to include the message definition files generated from PolySync’s ROS bridge in a new ROS node project. In order to start the tutorial, you will need the PolysyncLidarTranslator example.

1. Setup ROS workspace

To properly build this example with the generated PolySync messages in ROS, you will first need to set up your catkin workspace with the ros_bridge.

If you do not have a catkin workspace already, follow these steps to set one up.

http://wiki.ros.org/catkin/Tutorials/create_a_workspace

1.1 Create a buildable package

If you do have a workspace feel free to continue with the steps below. First source your ROS install and then create a package in the src directory of your catkin workspace as shown below.

$ cd ~/catkin_ws/src
$ source /opt/ros/indigo/setup.bash
$ catkin_create_pkg polysync_lidar_translator std_msgs roscpp tf

Copy “polysync_lidar_translator_node.cpp” to the src directory of the polysync_lidar_translator package you created.

$ cp PolySync-Core-CPP-Examples/PolysyncLidarTranslator/polysync_lidar_translator_node.cpp ~/catkin_ws/src/polysync_lidar_translator/src/

Make sure that in the “Build” section of the CMakeLists file for your PolysyncLidarTranslator project are uncommented. All other lines in the “Build” section should be commented.

## Declare a C++ executable
add_executable(polysync_lidar_translator_node src/polysync_lidar_translator_node.cpp)

## Add cmake target dependencies of the executable
## same as for the library above
add_dependencies(polysync_lidar_translator_node ${${PROJECT_NAME}_EXPORTED_TARGETS} ${catkin_EXPORTED_TARGETS})

## Specify libraries to link a library or executable target against
target_link_libraries(polysync_lidar_translator_node
   ${catkin_LIBRARIES}
 )

1.2 Generate the ROS bridge

Generate the ros_bridge files in the src folder of your catkin workspace and copy the ros_bridge project into the root directory of your catkin workspace, as shown below.

$ cd ~/catkin_ws/src
$ pdm-gen -r /usr/local/polysync/modules/sensor/sensor.idl /usr/local/polysync/modules/navigation/navigation.idl /usr/local/polysync/modules/control/control.idl /usr/local/polysync/modules/dtc/dtc.idl
$ mv pdm/ros_bridge ../
$ rm -rf pdm

2. Build and run the example

Source the ros_bridge, and then run catkin_make in the root directory of your catkin workspace.

$ cd ~/catkin_ws
$ source ros_bridge/setup.bash
$ catkin_make

Run ROS bridge node as shown below, remembering to source ros_bridge/setup.bash if using a different terminal.

$ cd ~/catkin_ws/ros_bridge/share/polysync_ros_bridge
$ ./BridgeNode -t ps_lidar_points_msg

3. Visualize data

Start up PolySync data generator in another terminal.

$ polysync-data-generator-c

Run the translator project you just built. The executable should be in the devel folder of your catkin workspace.

$ cd ~/catkin_ws/devel/lib/polysync_lidar_translator
$ ./polysync_lidar_translator_node

Open up rviz.

$ rosrun rviz rviz

In “Global Options” for the fixed frame select “world.” Then hit “add” and add a new “PointCloud2” visualization. For the topic of this new visualization select “/polysync/lidar_points.” You can change the style to “Points” to have a little more realistic looking LiDAR point cloud.

4. Further steps

This example is generalized to work with multiple LiDAR drivers working in PolySync. Each PointCloud2 message will contain the source GUID of the driver sending the ps_lidar_points_msg in its frame_id field. Thus, you can differentiate the incoming data as necessary. Similarly, the PolySync timestamp is copied over to the native PointCloud2 messages. The transform for each incoming message of a particular GUID can be manually modified to match what is in the PolySync SDF. This allows simple sensor translation so that the data is correctly kept relative to the vehicles center frame, as in PolySync. Finally, with a little work this example could be adapted to republish ps_radar_targets_msg to PointCloud2, allowing you to use all Radar drivers in PolySync with ROS. This example is intended to be very easy to use but very powerful in its potential applications.