• DepthAI ROS
  • Install from ros binaries
  • Install from source
  • DepthAI filters
  • Custom converters


luxonis/depthai-ros allows you to:
  • Use the cameras as classic RGBD sensors for your 3D vision needs.
  • Load Neural Networks and get the inference results straight from camera!
You can develop your ROS applications in following ways:
  • Use classes provided in depthai_bridge to construct your own driver (see stereo_inertial_node example on how to do that)
  • Use depthai_ros_driver package (currently available on ROS2 Humble and ROS Noetic) to get default experience (see ROS Driver on how)
Supported ROS versions:
  • Noetic
  • Humble
  • Iron

Install from ros binaries

Install depthai-ros. (Available for Noetic, Humble and Iron) sudo apt install ros-<distro>-depthai-ros

Install from source

For installing from source, please refer to this guide.

DepthAI filters

depthai_filters contains small composable node examples that show how to work with data from multiple topics. Available filters:
  • Detection2DOverlay - subscribes to /nn/detections and rgb/preview/image_raw topics. To see it in action, run ros2 launch depthai_filters example_det2d_overla.launch.py. Note here - If you see that detections misalign in the overlay, adjust rgb.i_preview_size parameter.
  • SegmentationOverlay, overlays semantic segmentation from /nn/image_raw on top of image from rgb/preview/image_raw, to see it in action, run ros2 launch depthai_filters example_seg_overlay.launch.py
  • WLS filter - stereo depth filter that smooths out overall depth image based on disparity data. It subscribes to stereo/image_raw and left/image raw topics. Parameters needed to enable it - left.i_publish_topic, stereo.i_output_disparity an example can be seen by running ros2 launch depthai_filters example_wls_filter.launch.py
  • SpatialBB - publishes bounding boxes as 3D line Markers based on spatial detections coming from driver node
  • FeatureTrackerOverlay - publishes Tracked Features overlay based on features and images coming from the driver
  • Features3D - uses depth image to republish features as 3D pointcloud

Custom converters

Users can write Custom converters and plug them in for bridge Publisher. If there a standard Message or usecase for which we have not provided a ros msg or converter feel free to create a issue or reach out to us on our discord community. We would be happy to add more.