rgbd_ros_to_lcm

ROS node to republish RGBD sensor data in LCM

This ROS package contains a ROS node that subscribes to RGB and depth images from an RGBD sensor (Kinect, RealSense, etc.) and republishes the data to an LCM channel. The output LCM message format matches the format of the openni2-camera-lcm driver and is compatible with the LabelFusion pipeline for generating ground truth labels of RGBD data.

For converting LCM messages to ROS, please refer to the lcm_to_ros package.

Installation

Dependencies

• LCM: Build Instructions
• PCL

Clone this repository into your catkin workspace:

cd ~/catkin_ws/src
git clone https://github.com/MobileManipulation/rgbd_ros_to_lcm.git
cd rgbd_ros_to_lcm
cd ~/catkin_ws
catkin_make

To generate C++ LCM message definitions for the image LCM types:

cd include
lcm-gen -x ../lcmtypes/*.lcm

Usage

The lcm_republisher ROS node subscribes to the output of an RGBD sensor driver ROS node, like the RealSense driver. First, launch the sensor driver. Then, to run the lcm_republisher node, use the example launch file provided:

roslaunch rgbd_ros_to_lcm lcm_republisher.launch

Parameters

Input Parameters

• subscribe_point_cloud: subscribe to an organized RGB point cloud topic instead of individually subscribing to RGB and depth topics
• cloud_topic: organized RGB point cloud topic
• rgb_topic: RGB image topic
• depth_topic: registered depth image topic

Note: If subscribing to RGB and depth image topics individually for use with LabelFusion, make sure the depth topic is registered into the RGB image frame to prevent alignment problems.

Output Parameters

• compress_rgb: use JPEG compression for RGB images
• compress_depth: use ZLIB compression for depth images
• output_lcm_channel: name of LCM channel for published output
• lcm_url: LCM URL

The default values of these parameters match the output of openni2-camera-lcm with both JPEG and ZLIB compression enabled. The data is published on the OPENNI_FRAME LCM channel.

Who do I talk to?

To report bugs or problems with the package, please open an issue on our issue tracker.