Marine Object Detection Using LiDAR on an Unmanned Surface Vehicle

Marine object detection plays a crucial role in various applications such as collision avoidance and autonomous navigation in maritime environments. While most existing datasets focus on 2D object detection, this research introduces a novel 3D object detection approach that relies exclusively on LiD...

Full description

Saved in:
Bibliographic Details
Main Authors: Yvan Eustache, Cedric Seguin, Antoine Pecout, Alexandre Foucher, Johann Laurent, Dominique Heller
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11075785/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849316734300848128
author Yvan Eustache
Cedric Seguin
Antoine Pecout
Alexandre Foucher
Johann Laurent
Dominique Heller
author_facet Yvan Eustache
Cedric Seguin
Antoine Pecout
Alexandre Foucher
Johann Laurent
Dominique Heller
author_sort Yvan Eustache
collection DOAJ
description Marine object detection plays a crucial role in various applications such as collision avoidance and autonomous navigation in maritime environments. While most existing datasets focus on 2D object detection, this research introduces a novel 3D object detection approach that relies exclusively on LiDAR (Light Detection And Ranging) data, specifically tailored for small Unmanned Surface Vehicles (USVs), where energy efficiency and computational constraints are key challenges. This study contributes a new point cloud dataset collected from a 2-meter autonomous USV and augmented through a hardware-in-the-loop simulation environment. The PointPillars network, chosen for its efficiency in processing LiDAR data, was trained and evaluated in this maritime context. A comparative analysis was also conducted between the proposed LiDAR-only method and a multimodal (LiDAR-camera) approach. The core innovation of this work is a step for late fusion strategy, where object detection is performed independently across sensors before integration. This results in a significantly less resource-intensive solution compared to early fusion methods. Consequently, the LiDAR-only approach highly suitable for deployment on compact, low-power autonomous surface drones, marking a step forward in practical and scalable marine perception systems.
format Article
id doaj-art-e438f7a3ec6f4044b1013d8b49e11144
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-e438f7a3ec6f4044b1013d8b49e111442025-08-20T03:51:29ZengIEEEIEEE Access2169-35362025-01-011312165812166910.1109/ACCESS.2025.358731511075785Marine Object Detection Using LiDAR on an Unmanned Surface VehicleYvan Eustache0https://orcid.org/0009-0007-6769-2581Cedric Seguin1Antoine Pecout2Alexandre Foucher3https://orcid.org/0009-0006-5723-9516Johann Laurent4https://orcid.org/0000-0001-7133-759XDominique Heller5Lab-STICC UMR 6285, Université Bretagne Sud, Lorient, FranceLab-STICC UMR 6285, Université Bretagne Sud, Lorient, FranceDiades Marine, Brest, FranceLab-STICC UMR 6285, Université Bretagne Sud, Lorient, FranceLab-STICC UMR 6285, Université Bretagne Sud, Lorient, FranceLab-STICC UMR 6285, Université Bretagne Sud, Lorient, FranceMarine object detection plays a crucial role in various applications such as collision avoidance and autonomous navigation in maritime environments. While most existing datasets focus on 2D object detection, this research introduces a novel 3D object detection approach that relies exclusively on LiDAR (Light Detection And Ranging) data, specifically tailored for small Unmanned Surface Vehicles (USVs), where energy efficiency and computational constraints are key challenges. This study contributes a new point cloud dataset collected from a 2-meter autonomous USV and augmented through a hardware-in-the-loop simulation environment. The PointPillars network, chosen for its efficiency in processing LiDAR data, was trained and evaluated in this maritime context. A comparative analysis was also conducted between the proposed LiDAR-only method and a multimodal (LiDAR-camera) approach. The core innovation of this work is a step for late fusion strategy, where object detection is performed independently across sensors before integration. This results in a significantly less resource-intensive solution compared to early fusion methods. Consequently, the LiDAR-only approach highly suitable for deployment on compact, low-power autonomous surface drones, marking a step forward in practical and scalable marine perception systems.https://ieeexplore.ieee.org/document/11075785/Deep learninghardware-in-the-loop simulatorLiDAR datasetmarine object detectionPointPillarsROS
spellingShingle Yvan Eustache
Cedric Seguin
Antoine Pecout
Alexandre Foucher
Johann Laurent
Dominique Heller
Marine Object Detection Using LiDAR on an Unmanned Surface Vehicle
IEEE Access
Deep learning
hardware-in-the-loop simulator
LiDAR dataset
marine object detection
PointPillars
ROS
title Marine Object Detection Using LiDAR on an Unmanned Surface Vehicle
title_full Marine Object Detection Using LiDAR on an Unmanned Surface Vehicle
title_fullStr Marine Object Detection Using LiDAR on an Unmanned Surface Vehicle
title_full_unstemmed Marine Object Detection Using LiDAR on an Unmanned Surface Vehicle
title_short Marine Object Detection Using LiDAR on an Unmanned Surface Vehicle
title_sort marine object detection using lidar on an unmanned surface vehicle
topic Deep learning
hardware-in-the-loop simulator
LiDAR dataset
marine object detection
PointPillars
ROS
url https://ieeexplore.ieee.org/document/11075785/
work_keys_str_mv AT yvaneustache marineobjectdetectionusinglidaronanunmannedsurfacevehicle
AT cedricseguin marineobjectdetectionusinglidaronanunmannedsurfacevehicle
AT antoinepecout marineobjectdetectionusinglidaronanunmannedsurfacevehicle
AT alexandrefoucher marineobjectdetectionusinglidaronanunmannedsurfacevehicle
AT johannlaurent marineobjectdetectionusinglidaronanunmannedsurfacevehicle
AT dominiqueheller marineobjectdetectionusinglidaronanunmannedsurfacevehicle