Accessible aerial robotics

摘要

This work demonstrates some of the computational capabilities of the inexpensive AR.Drone2 quadcopter. Fundamentally a platform that navigates in 3D, the drone offers a clean interface to 2D sensor data through its video stream. By using image-matching to bridge that interdimensional gap, this work demonstrates that the drone offers an accessible, compelling platform for research and independent projects at any institution. In the process, we demonstrate that accessible physical platforms, i.e., the drone, the Kinect, and their scaffolding software, ROS, now offer capabilities with much wider curricular and computational applications.