Date of Award

Fall 2024

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Electrical & Computer Engineering

Program/Concentration

Electrical and Computer Engineering

Committee Director

Masha Sosonkina

Committee Member

Yaohang Li

Committee Member

Lee Belfore

Abstract

Maritime autonomy, specifically the use of autonomous and semi-autonomous maritime vessels, is a key enabling technology supporting a set of diverse and critical research areas, including coastal and environmental resilience, assessment of waterway health, ecosystem/asset monitoring and maritime port security. Critical to the safe, efficient and reliable operation of an autonomous maritime vessel is its ability to perceive the external environment through onboard sensors. The main sensor utilized in this research is a LiDAR sensor. This sensor is able to generate point clouds of the surrounding environment, of which a machine learning model is used to label each point in the point cloud as belonging to a particular type of buoy. This classification of LiDAR scans is performed by using machine learning methods on data from a Unity Game Engine (herein referred to as Unity) simulation and real-life hand-labeled data from LiDAR scans. The Unity simulation data combined with labeled real-world maritime environment point cloud data were used for the training and testing of a PointNet-based neural network model. Fitting the PointNet-based model on the simulation and real-world data allowed for accurate classification of point clouds on the real-world data. Intersection Over Union (IoU) is the main metric used for this research which measures the overlap between predictions and ground truth labels. Different ratios of real to simulation data are experimented with, and the model that performed the best on the varying ratios of data had an IoU score of 0.84.

Rights

In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).

DOI

10.25777/vse7-jn66

ISBN

9798302855275

Download

Share

COinS