camera_pose_estimation: ar_pose_demo | ar_pose_ekf | ar_pose_world_convertor | rtt_ar_pose
Package Summary
ar_pose_ekf: Orocos component which fuses marker pose estimates from different cameras. The component uses a BFL (Bayesian Filter Library) Extended Kalman Filter to merge the different estimates. Input node: ar_pose::ARMarker message Output nodes: ar_pose::ARMarker message - visualization_msgs::Marker message Connection between the Orocos and ROS framework is done using the Orocos ROS integration stack.
- Author: Steven Bellens - steven.bellens@mech.kuleuven.be
- License: LGPLv2.1 - BSD
- Repository: kul-ros-pkg
- Source: git http://git.mech.kuleuven.be/robotics/camera_pose_estimation.git
It takes its estimates from the ar_pose_marker node (ARMarker message) using the ar_pose package and publishes fused estimated on the poseWithCovOut node (ARMarker message). The measurement covariance believe is updated according to the covariance contained in the ARMarker message. An additional ar_marker node is published containing a visualization_msgs::Marker message for rviz visualization.
Usage
Set up an ar_pose_marker node which gets pose estimates from your sensor. (demo launch files using prosilica cameras are available in the ar_pose_demo package)
- Start the ar_pose_ekf component
roscd ar_pose_ekf ./runekf.sh
Read out the fused estimated with e.g. rviz (demo vcg files are available in the ar_pose_demo package)