Date of Award
Summer 2025
Document Type
Open Access Thesis
Department
Mechanical Engineering
First Advisor
Yi Wang
Abstract
Achieving high reliability in small object detection is critical for advancing maritime and littoral intelligent navigation operations. Computer vision provides an undetectable form of perception that can be easily integrated into existing camera systems, offering a low-profile yet powerful solution for navigation and situational awareness in water-based environments. While deep learning techniques have significantly advanced the field of computer vision, the detection of small objects remains a persistent challenge. Limited pixel representation, the scarcity of large-scale small object datasets, and the reduced detail and indistinct features of small objects complicate detection efforts, especially in complex environments. These factors collectively hinder the ability of detection algorithms to accurately detect, localize, and classify small objects in maritime and littoral settings. To address these challenges, the IMSEL Maritime Dataset has been developed and labeled in accordance with UMAA guidelines, providing a dedicated resource aimed at improving the reliability of small object detection in maritime and littoral environments.
Rights
© 2025, Jacob Matthew Whisenant
Recommended Citation
Whisenant, J. M.(2025). Database Development for Reliable Small Object Detection in Maritime Environments. (Master's thesis). Retrieved from https://scholarcommons.sc.edu/etd/8537