Polarization-Enhanced Multi-Target Underwater Salient Object Detection

Salient object detection (SOD) plays a critical role in underwater exploration systems. Traditional SOD approaches encounter notable constraints in underwater image analysis, primarily stemming from light scattering and absorption effects induced by suspended particulate matter in complex underwater...

Full description

Saved in:
Bibliographic Details
Main Authors: Jiayi Song, Peikai Zhao, Jiangtao Li, Liming Zhu, Khian-Hooi Chew, Rui-Pin Chen
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Photonics
Subjects:
Online Access:https://www.mdpi.com/2304-6732/12/7/707
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849732769702215680
author Jiayi Song
Peikai Zhao
Jiangtao Li
Liming Zhu
Khian-Hooi Chew
Rui-Pin Chen
author_facet Jiayi Song
Peikai Zhao
Jiangtao Li
Liming Zhu
Khian-Hooi Chew
Rui-Pin Chen
author_sort Jiayi Song
collection DOAJ
description Salient object detection (SOD) plays a critical role in underwater exploration systems. Traditional SOD approaches encounter notable constraints in underwater image analysis, primarily stemming from light scattering and absorption effects induced by suspended particulate matter in complex underwater environments. In this work, we propose a deep learning-based multimodal method guided by multi-polarization parameters that integrates polarization de-scattering mechanisms with the powerful feature learning capability of neural networks to achieve adaptive multi-target SOD in an underwater turbid scattering environment. The proposed polarization-enhanced salient object detection network (PESODNet) employs a multi-polarization-parameter-guided, material-aware attention mechanism and a contrastive feature calibration unit, significantly enhancing its multi-material, multi-target detection capabilities in underwater scattering environments. The experimental results confirm that the proposed method achieves substantial performance improvements in multi-target underwater SOD tasks, outperforming state-of-the-art models of salient object detection in detection accuracy.
format Article
id doaj-art-6c8f13622b0042fb8ee799015d664588
institution DOAJ
issn 2304-6732
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Photonics
spelling doaj-art-6c8f13622b0042fb8ee799015d6645882025-08-20T03:08:13ZengMDPI AGPhotonics2304-67322025-07-0112770710.3390/photonics12070707Polarization-Enhanced Multi-Target Underwater Salient Object DetectionJiayi Song0Peikai Zhao1Jiangtao Li2Liming Zhu3Khian-Hooi Chew4Rui-Pin Chen5Key Laboratory of Optical Field Manipulation of Zhejiang Province, Department of Physics, Zhejiang Sci-Tech University, Hangzhou 310018, ChinaKey Laboratory of Optical Field Manipulation of Zhejiang Province, Department of Physics, Zhejiang Sci-Tech University, Hangzhou 310018, ChinaKey Laboratory of Optical Field Manipulation of Zhejiang Province, Department of Physics, Zhejiang Sci-Tech University, Hangzhou 310018, ChinaKey Laboratory of Optical Field Manipulation of Zhejiang Province, Department of Physics, Zhejiang Sci-Tech University, Hangzhou 310018, ChinaKey Laboratory of Optical Field Manipulation of Zhejiang Province, Department of Physics, Zhejiang Sci-Tech University, Hangzhou 310018, ChinaKey Laboratory of Optical Field Manipulation of Zhejiang Province, Department of Physics, Zhejiang Sci-Tech University, Hangzhou 310018, ChinaSalient object detection (SOD) plays a critical role in underwater exploration systems. Traditional SOD approaches encounter notable constraints in underwater image analysis, primarily stemming from light scattering and absorption effects induced by suspended particulate matter in complex underwater environments. In this work, we propose a deep learning-based multimodal method guided by multi-polarization parameters that integrates polarization de-scattering mechanisms with the powerful feature learning capability of neural networks to achieve adaptive multi-target SOD in an underwater turbid scattering environment. The proposed polarization-enhanced salient object detection network (PESODNet) employs a multi-polarization-parameter-guided, material-aware attention mechanism and a contrastive feature calibration unit, significantly enhancing its multi-material, multi-target detection capabilities in underwater scattering environments. The experimental results confirm that the proposed method achieves substantial performance improvements in multi-target underwater SOD tasks, outperforming state-of-the-art models of salient object detection in detection accuracy.https://www.mdpi.com/2304-6732/12/7/707deep learningpolarization imagingunderwater salient object detection
spellingShingle Jiayi Song
Peikai Zhao
Jiangtao Li
Liming Zhu
Khian-Hooi Chew
Rui-Pin Chen
Polarization-Enhanced Multi-Target Underwater Salient Object Detection
Photonics
deep learning
polarization imaging
underwater salient object detection
title Polarization-Enhanced Multi-Target Underwater Salient Object Detection
title_full Polarization-Enhanced Multi-Target Underwater Salient Object Detection
title_fullStr Polarization-Enhanced Multi-Target Underwater Salient Object Detection
title_full_unstemmed Polarization-Enhanced Multi-Target Underwater Salient Object Detection
title_short Polarization-Enhanced Multi-Target Underwater Salient Object Detection
title_sort polarization enhanced multi target underwater salient object detection
topic deep learning
polarization imaging
underwater salient object detection
url https://www.mdpi.com/2304-6732/12/7/707
work_keys_str_mv AT jiayisong polarizationenhancedmultitargetunderwatersalientobjectdetection
AT peikaizhao polarizationenhancedmultitargetunderwatersalientobjectdetection
AT jiangtaoli polarizationenhancedmultitargetunderwatersalientobjectdetection
AT limingzhu polarizationenhancedmultitargetunderwatersalientobjectdetection
AT khianhooichew polarizationenhancedmultitargetunderwatersalientobjectdetection
AT ruipinchen polarizationenhancedmultitargetunderwatersalientobjectdetection