Deep Metric Learning for Fine-Grained Ship Classification in SAR Images with Sidelobe Interference

The interference of sidelobe often causes different targets to exhibit similar features, diminishing fine-grained classification accuracy. This effect is particularly pronounced when the available data are limited. To address the aforementioned issues, a novel classification framework for sidelobe-a...

Full description

Saved in:
Bibliographic Details
Main Authors: Haibin Zhu, Yaxin Mu, Wupeng Xie, Kang Xing, Bin Tan, Yashi Zhou, Zhongde Yu, Zhiying Cui, Chuang Zhang, Xin Liu, Zhenghuan Xia
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/17/11/1835
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850158961477550080
author Haibin Zhu
Yaxin Mu
Wupeng Xie
Kang Xing
Bin Tan
Yashi Zhou
Zhongde Yu
Zhiying Cui
Chuang Zhang
Xin Liu
Zhenghuan Xia
author_facet Haibin Zhu
Yaxin Mu
Wupeng Xie
Kang Xing
Bin Tan
Yashi Zhou
Zhongde Yu
Zhiying Cui
Chuang Zhang
Xin Liu
Zhenghuan Xia
author_sort Haibin Zhu
collection DOAJ
description The interference of sidelobe often causes different targets to exhibit similar features, diminishing fine-grained classification accuracy. This effect is particularly pronounced when the available data are limited. To address the aforementioned issues, a novel classification framework for sidelobe-affected SAR imagery is proposed. First, a method based on maximum median filtering is adopted to remove sidelobe by exploiting local grayscale differences between the target and sidelobe, constructing a high-quality SAR dataset. Second, a deep metric learning network is constructed for fine-grained classification. To enhance the classification performance of the network on limited samples, a feature extraction module integrating a lightweight attention mechanism is designed to extract discriminative features. Then, a hybrid loss function is proposed to strengthen intra-class correlation and inter-class separability. Experimental results based on the FUSAR-Ship dataset demonstrate that the method exhibits excellent sidelobe suppression performance. Furthermore, the proposed framework achieves an accuracy of 84.18% across five ship target classification categories, outperforming the existing methods, significantly enhancing the classification performance in the context of sidelobe interference and limited datasets.
format Article
id doaj-art-db206d2d30f7480eaf4a0f24972a7442
institution OA Journals
issn 2072-4292
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-db206d2d30f7480eaf4a0f24972a74422025-08-20T02:23:44ZengMDPI AGRemote Sensing2072-42922025-05-011711183510.3390/rs17111835Deep Metric Learning for Fine-Grained Ship Classification in SAR Images with Sidelobe InterferenceHaibin Zhu0Yaxin Mu1Wupeng Xie2Kang Xing3Bin Tan4Yashi Zhou5Zhongde Yu6Zhiying Cui7Chuang Zhang8Xin Liu9Zhenghuan Xia10School of Information and Communication Engineering, Beijing Information Science and Technology University, Beijing 100101, ChinaSchool of Information and Communication Engineering, Beijing Information Science and Technology University, Beijing 100101, ChinaArtificial Intelligence Institute of China Electronics Technology Group Corporation, Beijing 100041, ChinaBeijing Institute of Satellite Information Engineering, Beijing 100095, ChinaBeijing Institute of Satellite Information Engineering, Beijing 100095, ChinaChina Academy of Space Technology, Beijing 100094, ChinaBeijing Institute of Satellite Information Engineering, Beijing 100095, ChinaBeijing Institute of Satellite Information Engineering, Beijing 100095, ChinaBeijing Institute of Satellite Information Engineering, Beijing 100095, ChinaBeijing Institute of Satellite Information Engineering, Beijing 100095, ChinaBeijing Institute of Satellite Information Engineering, Beijing 100095, ChinaThe interference of sidelobe often causes different targets to exhibit similar features, diminishing fine-grained classification accuracy. This effect is particularly pronounced when the available data are limited. To address the aforementioned issues, a novel classification framework for sidelobe-affected SAR imagery is proposed. First, a method based on maximum median filtering is adopted to remove sidelobe by exploiting local grayscale differences between the target and sidelobe, constructing a high-quality SAR dataset. Second, a deep metric learning network is constructed for fine-grained classification. To enhance the classification performance of the network on limited samples, a feature extraction module integrating a lightweight attention mechanism is designed to extract discriminative features. Then, a hybrid loss function is proposed to strengthen intra-class correlation and inter-class separability. Experimental results based on the FUSAR-Ship dataset demonstrate that the method exhibits excellent sidelobe suppression performance. Furthermore, the proposed framework achieves an accuracy of 84.18% across five ship target classification categories, outperforming the existing methods, significantly enhancing the classification performance in the context of sidelobe interference and limited datasets.https://www.mdpi.com/2072-4292/17/11/1835synthetic aperture radarfine-grained ship classificationsidelobe effectmaximum median filterdeep metric learning
spellingShingle Haibin Zhu
Yaxin Mu
Wupeng Xie
Kang Xing
Bin Tan
Yashi Zhou
Zhongde Yu
Zhiying Cui
Chuang Zhang
Xin Liu
Zhenghuan Xia
Deep Metric Learning for Fine-Grained Ship Classification in SAR Images with Sidelobe Interference
Remote Sensing
synthetic aperture radar
fine-grained ship classification
sidelobe effect
maximum median filter
deep metric learning
title Deep Metric Learning for Fine-Grained Ship Classification in SAR Images with Sidelobe Interference
title_full Deep Metric Learning for Fine-Grained Ship Classification in SAR Images with Sidelobe Interference
title_fullStr Deep Metric Learning for Fine-Grained Ship Classification in SAR Images with Sidelobe Interference
title_full_unstemmed Deep Metric Learning for Fine-Grained Ship Classification in SAR Images with Sidelobe Interference
title_short Deep Metric Learning for Fine-Grained Ship Classification in SAR Images with Sidelobe Interference
title_sort deep metric learning for fine grained ship classification in sar images with sidelobe interference
topic synthetic aperture radar
fine-grained ship classification
sidelobe effect
maximum median filter
deep metric learning
url https://www.mdpi.com/2072-4292/17/11/1835
work_keys_str_mv AT haibinzhu deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT yaxinmu deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT wupengxie deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT kangxing deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT bintan deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT yashizhou deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT zhongdeyu deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT zhiyingcui deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT chuangzhang deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT xinliu deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference
AT zhenghuanxia deepmetriclearningforfinegrainedshipclassificationinsarimageswithsidelobeinterference