wajeehulhassanvii / computer-vision-ardrone

ROS library for an AR drone to follow an Aruco tag in-flight

Computer Vision (Tag Tracking) AR Drone

This ROS package utilizes OpenCV and ARDrone_Autonomy on ROS to have a Parrot AR Drone follow an augmented reality tag.

Check out a short demo video here

Prerequisites

• ROS (Kinetic) - Provides the basic I/O and infrastructure for the AR Drone
• OpenCV - Open Source computer vision library to process images from the Drone
• ArUco - simple-to-use fast augmented reality tage recognition library
• ardrone_autonomy - ROS driver for the AR Drone 1.0 and 2.0
• aruco_ros - ROS wrapper for the ArUco library, publishes pose information to ROS topics

ROS Workspace Setup

After installing ROS, OpenCV, ArUco and ardrone_autonomy, initialize a new catkin workspace. In the src folder clone the aruco_ros library and catkin_make. then clone this repo to the src folder and make again.

Startup and Controls

There is probably a more organized way to do this but I ran 3 different terminals to run the commands, the image_view window from the drone must be in focus for the key inputs to work correctly.

• bring up the AR Drone and connect to the wireless network
• In the first terminal run roslaunch ardrone_autonomy ardrone.launch to bring up the ardrone topics
• In the second terminal run roslaunch aruco_ros single.launch markerId:=100 markerSize:=0.168 to bring up the tag recognition
• In a third terminal run rosrun ardrone_tag_follow ardrone_tag_follow.py to bring up the follow control
• press t to takeoff and space to land

NOTE: The drone will follow the last known position of the tag, if it's within the deadzone the drone will continue to hover.

Notes

• to view aruco debug info run rosrun image_view image_view image:=/aruco_single/result
• to switch camera on the ardrone run rosservice call /ardrone/togglecam
• the tags folder has some simple AR tags that can be printed to use with aruco libraries