tf_mapping

Overview

This package extends aruco_mapping with many features. You can use it to first create a map of aruco markers and then to navigate a simple camera (without depth information) in this map.

Author: Markus Lamprecht
Maintainer: Markus Lamprecht, 2f4yor@gmail.com

Video:

Installation

Dependencies

This software is built on the Robotic Operating System (ROS), which needs to be installed first. Additionally, this package depends on following software:

• alfons_msgs
• aruco_detector_ocv
• simple_kf
• rviz_pics
• opencv

Install the above packages first!

Building

In order to install this package, clone the latest version from this repository into your catkin workspace and compile the package using catkin_tools :

mkdir -p catkin_ws/src
cd catkin_ws/src/
git clone git@github.com:CesMak/tf_mapping.git 
cd ..
catkin init 
catkin build
source devel/setup.bash

Basic Usage

1. First step is to create a map:

       roslaunch tf_mapping start_tf_mapping.launch
   

   Store the tf's and the images using:

       rosrun tf_mapping tf_create_map.py
   

   • Adjust full_hd resolution (or the highes resolution for creating the map) in tf_mapping/launch/start_tf_mapping.launch
   • Point the camera such that only the aruco marker with id=0 is detected (This will be your initial coordinate system).
   • The transformations of all other detected aruco markers will be relative to this initial system.
   • Point the camera to another aruco marker e.g. with id=1
   • When moving the camera consider that always 2 markers are visible on every image.

2. Navigate your camera now in this map

   • Adjust to use now lower resolution e.g. 640x480 (to be realtime capable) in tf_mapping/launch/start_tf_navigate.launch

       roslaunch tf_mapping start_tf_navigate.launch
   

Main Launch file

roslaunch tf_mapping start_tf_navigate.launch

License BSD

If you want to use this package please contact: me.