A Deep Learning-Based Automatic Object Detection Method for Autonomous Driving Ships

<p>An important feature of an Autonomous Surface Vehicles (ASV) is its capability of automatic object detection to avoid collisions, obstacles and navigate on their own.</p> <p>Deep learning has made some significant headway in solving fundamental challenges associated with object detection and computer vision. With tremendous demand and advancement in the technologies associated with ASVs, a growing interest in applying deep learning techniques in handling challenges pertaining to autonomous ship driving has substantially increased over the years.</p> <p>In this thesis, we study, design, and implement an object recognition framework that detects and recognizes objects found in the sea. We first curated a Sea-object Image Dataset (SID) specifically for this project. Then, by utilizing a pre-trained RetinaNet model on a large-scale object detection dataset named Microsoft COCO, we further fine-tune it on our SID dataset. We focused on sea objects that may potentially cause collisions or other types of maritime accidents. Our final model can effectively detect various types of floating or surrounding objects and classify them into one of the ten predefined significant classes, which are buoy, ship, island, pier, person, waves, rocks, buildings, lighthouse, and fish. Experimental results have demonstrated its good performance.</p>

Name thesis:degree_name

Master of Science (MS)

Level thesis:degree_level

Masters Thesis

Year dc:date.available

2021

Author dc:creator

• Atawodi, Ojonoka Erika

Contributors dc:contributor

• Dr Bo Li
• Dr Lina Pu
• Dr Chaoyang Zhang

• object detection
• deep learning
• autonomous driving ships
• RetinaNet
• Artificial Intelligence and Robotics
• Computational Engineering
• Computer Engineering
• Computer Sciences
• Navigation, Guidance, Control, and Dynamics
• Other Computer Engineering
• Other Computer Sciences

Repository record dc:identifier

https://aquila.usm.edu/masters_theses/813

OAI identifier oai:identifier

oai:aquila.usm.edu:masters_theses-1870