Agile vision-based tracking: Hardware platform, simulator, and performance evaluation
Yang, Benjamin Cong
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/125736
Description
Title
Agile vision-based tracking: Hardware platform, simulator, and performance evaluation
Author(s)
Yang, Benjamin Cong
Issue Date
2024-07-19
Director of Research (if dissertation) or Advisor (if thesis)
Unmanned Aerial Vehicles (UAVs) have been widely used in many applications and, the design and control of autonomous UAVs have attracted attention from researchers. Recent advancements in computer vision techniques have further enabled UAVs to perform complex tasks using relatively inexpensive cameras. Inspired by developments, this thesis explores the topic of vision-based target tracking for multi-rotor UAVs. Specifically, we address the challenge of a chaser drone tracking a leader drone that follows complex trajectories using only vision feedback. The drone tracking problem we are solving is central to critical applications, such as search and rescue, environmental monitoring, and coordinated flight operations. The complexity of this problem arises not only from the need to accurately identify and follow the leader drone using a vision pipeline but also from the necessity to maintain robustness despite environmental variations and the dynamic movements of the leader drone. The chaser drone must be capable of responding effectively when visual contact with the leader is temporarily lost.
In this thesis, we present the design an end-to-end autonomous pipeline for performing this drone tracking task. We present a control strategy that utilizes trajectory predictions generated by a Monte Carlo Prediction algorithm for the chaser drone when visual information about the leader drone is unavailable. The proposed pipeline is initially tested in a simulated environment. We further evaluate its performance on actual hardware, designed and built in house, based on Agilicious open-source autonomous quadcopter system. Under the space constraints of the indoor flying arena we deploy our experiments in, the result of experiments demonstrate that the pipeline we developed indeed allow the chaser drone to track the leader drone as long as target detection is successful. Moreover, even when the chaser drone temporarily loses visual contact in the flying arena, our control strategy allows it to regain visual contact and continue tracking effectively.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
