/Integrating-aruco_mapping-with-hector_SLAM

Localization using aruco markers on the map built by hector_slam with aruco markers inserter in it.

MIT LicenseMIT

Overview

In this project localization using aruco markers has been implemeted. Firstly, a binary grid occupancy map of the arena is created using the hector SLAM algorithm. Aruco markers are already present in the arena and are mapped in real-time with hector_slam on the same map. Camera pose(6-DOF) from these aruco markers are then used for localization and path-planning. The amcl package of navigation_stack uses the camera pose generated by laser scanner to generate the transform between /map and /odom. Instead of this camera pose, camera pose from aruco markers is used to generate the above transform.

The demonstration is divided into two stages. Both stage 1 and 2 have been implemented in the form of Gazebo Simulation and real-life implementation on the actual lizi bot. Stage1 : Mapping using hector_slam with realtime insertion of aruco markers in the map. Stage2 : The above map is loaded. Path-planning and localization has been done with using aruco markers and odometry. Aruco_markers estimate pose which is used to correct the drift in odometry calculation.

System Specifications

  • ros-indigo
  • ubuntu 14.04
  • lizi robot

Packages used and modified

New Packages

STAGE 1 Commands for real-life mapping

roslaunch lizi_mapping hector.launch

roslaunch aruco_mapping aruco_mapping.launch

rosrun aruco_tf_lizi lizi_broadcaster

STAGE 2 Commands for Gazebo Simulation

roslaunch aruco_mapping aruco_mapping.launch

roslaunch aruco_turtle_simulation nav_turtle.launch

rosrun aruco_tf_lizi read_bag

rosrun aruco_tf_lizi view_markers

Authors-