Autonomous navigation and tracking of dynamic surface targets on-board a computationally impoverished aerial vehicle
Year: 2011
Type: dissertation
Abstract: This thesis describes the development of an independent, on-board visual servoing system which allows a computationally impoverished aerial vehicle to autonomously identify and track a dynamic surface target.Image segmentation and target tracking algorithms are developed for the specific task of monitoring whales at sea.The computer vision algorithms' estimates prove to be accurate enough for quadrotor stabilization while being computationally fast enough to be processed on-board the platform.This differs from current techniques which require off-board processing of images for vehicle localization and control.The vision algorithm is evaluated on video This thesis would not have been possible without the support of a large group of friends, family, and colleagues.I would primarily like to thank my advisor, Daniela Rus, whose seemingly unlimited energy kept me motivated and excited throughout this research.Her wide ranging interests in all aspects of robotics allowed me unique opportunities to apply this research to a multitude of related projects.I would also like to thank my computer vision mentor, Peter Corke.Peter's wide range and experience in visual servoing helped guide key design choices in vision as well as control.I am especially grateful for the time spent at his lab at QUT in Australia.I am especially grateful to my fellow Distributed Robotics Lab colleagues who took time from their own research to take critical looks at mine.Daniel Soltero was invaluable as a second hand when running experiments.Kyle Gilpin and Marek Doniec provided very valuable hardware advice that saved countless hours of debugging.I would also like to thank Abe Bacharach of the Robust Robotics Group at MIT for access to several key software components for this research and his extensive quadrotor control knowledge.I am especially grateful to Brian Julian for his experience in dealing with the quadrotor platform as well as his large knowledge of control and estimation theory.I wish all my lab mates the most of luck in their future endeavors.
Author: William Clayton Selby
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1521065851