robokudo.vis.ros_visualizer

ROS-based visualization for RoboKudo pipelines.

This module provides ROS-based visualization capabilities for RoboKudo pipelines. It handles:

  • Image publishing to ROS topics

  • Single and multi-view visualization

  • Pipeline state visualization

  • Dynamic topic management

  • Image format conversion

Dependencies

  • rospy for ROS integration

  • cv2 for image manipulation

  • numpy for array operations

  • cv_bridge for ROS/OpenCV conversion

  • robokudo.annotators for annotator access

  • robokudo.vis.visualizer for base visualization interface

See Also

Classes

SharedROSVisualizer

A single-view ROS Image Publisher. It publishes the active annotator from the SharedState.

AllAnnotatorROSVisualizer

A ROS Image Publisher that publishes the output of all images in the given Pipeline.

Module Contents

class robokudo.vis.ros_visualizer.SharedROSVisualizer(*args, **kwargs)

Bases: robokudo.vis.visualizer.Visualizer, robokudo.vis.visualizer.Visualizer.Observer, rclpy.node.Node

A single-view ROS Image Publisher. It publishes the active annotator from the SharedState.

Variables:

  • shared_visualizer_state – Shared state object for coordinating between visualizers

  • ros_image_publisher – Publisher for the image topic

  • ros_image_cv_bridge – Bridge for converting between ROS and OpenCV image formats

  • update_output – Flag indicating if display needs updating

ros_image_publisher
ros_image_cv_bridge
tick()

Update the visualization display.

This method:

  • Gets current annotator outputs

  • Updates display if needed

  • Renders annotator name overlay

  • Publishes image to ROS topic

notify(observable, *args, **kwargs)

Handle notification of state changes.

Parameters:

observable (object) – The object that sent the notification

class robokudo.vis.ros_visualizer.AllAnnotatorROSVisualizer(*args, **kwargs)

Bases: robokudo.vis.visualizer.Visualizer, rclpy.node.Node

A ROS Image Publisher that publishes the output of all images in the given Pipeline.

This class provides visualization of all annotator outputs in a pipeline through separate ROS image topics. It supports:

  • Multiple image topic publishing

  • Dynamic topic creation

  • Image format conversion

  • Per-annotator output streams

Variables:

  • ros_image_publishers – Mapping of annotator names to ROS publishers

  • ros_image_cv_bridge – Bridge for converting between ROS and OpenCV image formats

  • update_output – Flag indicating if display needs updating

ros_image_publishers
ros_image_cv_bridge
update_ros_image_publishers()

Update ROS publishers for all annotators.

This method:

  • Gets current list of annotators

  • Creates publishers for new annotators

  • Maintains existing publishers

TODO: Consider removing publishers for non-existing annotators. May not be worth the cost if re-instantiation is frequent due to changing annotator lists.

tick()

Update all visualization displays.

This method:

  • Gets current annotator outputs

  • Updates publishers if needed

  • Publishes all annotator images to ROS topics

notify(observable, *args, **kwargs)

Handle notification of state changes.

Parameters:

observable (object) – The object that sent the notification