Store the latest detection on self so another callback can use it.

self.target_visible = False
self.target_position = 'UNKNOWN'

# Inside image_callback, after processing pixels:
self.target_visible = visible
self.target_position = position

Camera callback → node state → other callbacks.

On a real ROS 2 machine, use an ament_python package for Python or an ament_cmake package with rclcpp for C++. Declare dependencies in package.xml.

For C++, CMakeLists.txt uses find_package, add_executable, ament_target_dependencies and install(TARGETS ...) to build and install the node. Build with colcon build, source the workspace setup, then launch with ros2 run.

KineNest provides educational compatibility headers in the browser. The kinenest::report_* helpers record learning evidence; they are not native ROS 2 APIs. No package build runs here.

Real environment