Boundary Enhancement-Driven Accurate Semantic Segmentation Networks for Unmanned Surface Vessels in Complex Marine Environments

Semantic segmentation of complex marine environments based on images is crucial for the autonomous navigation of unmanned surface vessels (USVs). However, existing semantic segmentation methods mainly categorize images into three coarse categories: sea surface, obstacles, and sky, with limited atten...

Full description

Saved in:
Bibliographic Details
Published in:IEEE sensors journal 2024-08, Vol.24 (15), p.24972-24987
Main Authors: Zhang, Liye, Sun, Xiaoyu, Li, Zhongzheng, Kong, Dong, Liu, Jigang, Ni, Peizhou
Format: Article
Language:English
Subjects:
Accuracy
Autonomous navigation
Autonomous robots
Boundary enhancement
complex maritime environment
Datasets
Feature extraction
Image enhancement
Image segmentation
Marine environment
Modules
Obstacle avoidance
Sea surface
Sea vessels
Semantic segmentation
Semantics
Sensors
unmanned surface vessel (USV)
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Semantic segmentation of complex marine environments based on images is crucial for the autonomous navigation of unmanned surface vessels (USVs). However, existing semantic segmentation methods mainly categorize images into three coarse categories: sea surface, obstacles, and sky, with limited attention to the boundaries between significant elements. Moreover, the dynamic changes in marine environments often affect these methods, resulting in shortcomings such as blurry boundary segmentation. To address these issues, this article proposes a boundary enhancement-driven semantic segmentation network tailored for complex marine environments, focusing primarily on the accuracy of interclass boundaries. Specifically, 1) a boundary extraction module is proposed to extract multiscale boundary features from the backbone network, which are fused via continuous boundary attention streams (BASs) and supervised with boundary loss. 2) A boundary enhancing module (BEM) is introduced, wherein the connection between channel features is strengthened using the astous convolutional pyramid module and channel attention (CA) module, enhancing the perception of contextual boundary information and improving the accuracy of category boundaries in the segmentation results. Comprehensive comparative experiments were conducted on the MaSTr1325 dataset and MID dataset, and the proposed method was evaluated on the MODD2 dataset using the MODS evaluation method. Results demonstrate that our approach, BEMSNet, achieves clearer and more accurate boundary segmentation compared to other networks.
ISSN:1530-437X
1558-1748
DOI:10.1109/JSEN.2024.3409756