面向ISAR舰船目标的多域交互识别网络

doi:10.12305/j.issn.1001-506X.2026.11.08

系统工程与电子技术 ›› 2026, Vol. 48 ›› Issue (11): 3761-3769.doi: 10.12305/j.issn.1001-506X.2026.11.08

面向ISAR舰船目标的多域交互识别网络

李晓迪(), 李泓序(), 宿富林, 胡悦

哈尔滨工业大学电子与信息工程学院，黑龙江哈尔滨 150001

收稿日期:2025-06-24 修回日期:2025-11-06 接受日期:2025-11-26 出版日期:2026-10-25 发布日期:2026-03-18
通讯作者: 李泓序 E-mail:22B305004@stu.hit.edu.cn;20B905041@stu.hit.edu.cn
作者简介:李晓迪（1994—），女，博士研究生，主要研究方向为信号处理、人工智能
宿富林（1964—），男，教授，博士，主要研究方向为雷达成像、目标识别
胡　悦（1986—），女，教授，博士，主要研究方向为信号处理、人工智能
基金资助:
国家自然科学基金(62371167)；黑龙江省自然科学基金(YQ2021F005)资助课题

Multi-domain interactive recognition network for ISAR ship target

Xiaodi Li(), Hongxu Li(), Fulin Su, Yue Hu

School of Electronics and Information Engineering，Harbin Institute of Technology，Harbin 150001，China

Received:2025-06-24 Revised:2025-11-06 Accepted:2025-11-26 Online:2026-10-25 Published:2026-03-18
Contact: Hongxu Li E-mail:22B305004@stu.hit.edu.cn;20B905041@stu.hit.edu.cn

摘要/Abstract

摘要：

针对现有雷达目标识别方法通常仅单独处理高分辨率距离像（high-resolution range profiles，HRRP）和逆合成孔径雷达（inverse synthetic aperture radar，ISAR）图像特征，未能充分挖掘其互补的结构与运动信息的问题，提出一种联合利用HRRP与ISAR特征的多域交互识别网络（multi-domain interactive recognition network，MIR-Net），用于ISAR舰船目标识别。首先，MIR-Net采用对称尺度HRRP编码器和图像编码器，分别提取脉冲序列和图像的多尺度特征，并使用图卷积网络进一步建模其时空依赖。然后，构建多域特征交互融合模块，将两个域特征映射至共享潜在空间，并利用双分支交叉注意力机制实现动态融合。最后，融合HRRP特征、图像特征与联合表示进行目标的分类。在仿真数据集和实测数据集上的实验结果证明，所提方法能够提高ISAR目标识别准确率，具有有效性与鲁棒性。

关键词: 高分辨率距离像, 逆合成孔径雷达, 雷达目标识别, 多域特征交互

Abstract:

To address the problem that existing radar target recognition methods typically only process high-resolution range profiles （HRRPs） and inverse synthetic aperture radar （ISAR） image features separately, failing to fully exploit their complementary structural and motion information, a multi-domain interactive recognition network （MIR-Net） that jointly utilizes HRRP and ISAR features for ISAR ship target recognition is proposed. Firstly, MIR-Net employs a symmetric-scale HRRP encoder and an image encoder to extract multi-scale features from pulse sequences and images respectively, while a graph convolutional network is used to further capture spatiotemporal dependencies. Subsequently, a multi-domain feature interaction fusion module is introduced to project both domains into a shared latent space and dynamically fuse them using a dual-branch cross-attention mechanism. Finally, HRRP features, image features, and the fused representation are combined for classification. Experimental results on both simulated and measured datasets demonstrate that the proposed method can improve the accuracy of ISAR target recognition, and have effectiveness and robustness.

Key words: high-resolution range profile (HRRP), inverse synthetic aperture radar (ISAR), radar target recognition, multi-domain feature interaction

中图分类号:

TN 957.51

李晓迪, 李泓序, 宿富林, 胡悦. 面向ISAR舰船目标的多域交互识别网络[J]. 系统工程与电子技术, 2026, 48(11): 3761-3769.

Xiaodi Li, Hongxu Li, Fulin Su, Yue Hu. Multi-domain interactive recognition network for ISAR ship target[J]. Systems Engineering and Electronics, 2026, 48(11): 3761-3769.

图/表 15

图1

图2

图3

表1

图4

表2

表3

表4

图5

图6

图7

图8

图9

表5

表6

参考文献 30

1	Huang P H, Xia X G, Zhan M, et al. ISAR imaging of a maneuvering target based on parameter estimation of multicomponent cubic phase signals[J]. IEEE Trans. on Geoscience and Remote Sensing, 2021, 60, 5103918.
2	李开明, 张袁鹏, 罗迎, 等. 弹道导弹雷达目标识别研究进展[J]. 系统工程与电子技术, 2025, 47 (9): 2870. doi: 10.12305/j.issn.1001-506X.2025.09.10
3	王国庆, 张云龙, 但波, 等. 基于高效局部注意力的小样本HRRP目标识别网络[J]. 雷达科学与技术, 2025, 23 (6): 603.
4	王彩云, 贾一帆, 李晓飞, 等. 基于改进鲸鱼算法和BiGRU的弹道目标HRRP识别方法[J]. 系统工程与电子技术, 2025, 47 (6): 1824. doi: 10.12305/j.issn.1001-506X.2025.06.11
5	Liu X D, Wang L, Bai X R. End-to-end radar HRRP target recognition based on integrated denoising and recognition network[J]. Remote Sensing, 2022, 14 (20): 5254. doi: 10.3390/rs14205254
6	Li W T, Tian B, Ma J, et al. OSFSM: a systematic open set framework for radar automatic target recognition using HRRP[J]. IEEE Trans. on Aerospace and Electronic Systems, 2025, 61 (5): 12350. doi: 10.1109/TAES.2025.3575045
7	Dong G G, Liu H W. Target recognition in ISAR image via range profile perturbation imaging[J]. IEEE Trans. on Geoscience and Remote Sensing, 2024, 62, 3502695. doi: 10.1109/TGRS.2024.3502695
8	Ni H, Jiang Y C, Chen R D, et al. STRE-diffusion: few-shot spaceborne ISAR target recognition via spatial-time relationship estimation and improved diffusion model[J]. IEEE Trans. on Aerospace and Electronic Systems, 2025, 61 (6): 17906.
9	Ni P S, Liu Y Y, Pei H, et al. CLISAR-NET: a deformation-robust isar image classification network using contrastive learning[J]. Remote Sensing, 2022, 15 (1): 33. doi: 10.3390/rs15010033
10	Wu C X, Sun J M, Cheng Q, et al. Radar target recognition based on data-level fusion of radar multi-dimensional information[C]//IET Conference Proceedings, 2023.
11	Tian X P, Bai X R, Zhou F. Recognition of micro-motion space targets based on attention-augmented cross-domain feature fusion recognition network[J]. IEEE Trans. on Geoscience and Remote Sensing, 2023, 61, 5104909.
12	Deng J, Su F L. SDRNet: a deep fusion network for ISAR ship target recognition based on feature separation and weighted decision[J]. Remote Sensing, 2024, 16 (11): 1920. doi: 10.3390/rs16111920
13	Liu Q, Zhang X Y, Liu Y X. A prior-knowledge-guided neural network based on supervised contrastive learning for radar HRRP recognition[J]. IEEE Trans. on Aerospace and Electronic Systems, 2024, 60 (3): 2854. doi: 10.1109/TAES.2024.3353718
14	Chen L F, Sun X, Pan Z L, et al. HRRPGraphNet: make HRRPs to be graphs for efficient target recognition[J]. Electronics Letters, 2024, 60 (22): e70088. doi: 10.1049/ell2.70088
15	Zhang X W, Wei Y, Wang W Y. Patch-wise autoencoder based on transformer for radar high-resolution range profile target recognition[J]. IEEE Sensors Journal, 2023, 23 (23): 29406. doi: 10.1109/JSEN.2023.3327552
16	Gao S K, Bie B W, Wang H K. Radar high-resolution range profile target recognition method based on structure-aware network[J]. IEEE Sensors Journal, 2024, 24 (20): 32660. doi: 10.1109/JSEN.2024.3449571
17	Zhou X N, Bai X R, Wang L, et al. Robust ISAR target recognition based on ADRISAR-Net[J]. IEEE Trans. on Aerospace and Electronic Systems, 2022, 58 (6): 5494. doi: 10.1109/TAES.2022.3174826
18	Yuan H X, Li H B, Zhang Y, et al. Complex-valued multiscale vision transformer on space target recognition by ISAR image sequence[J]. IEEE Geoscience and Remote Sensing Letters, 2024, 21, 4008305.
19	Deng J, Su F L. Deep hybrid fusion network for inverse synthetic aperture radar ship target recognition using multi-domain high-resolution range profile data[J]. Remote Sensing, 2024, 16 (19): 3701.
20	Zhou Y B, Liu S, Gao H W, et al. A dual-polarization feature fusion network for radar automatic target recognition based on HRRP sequences[J]. IEEE Trans. on Aerospace and Electronic Systems, 2025, 61 (5): 14642.
21	徐鹏政. 基于HRRP与ISAR图像融合的舰船目标识别研究[D]. 哈尔滨工业大学, 2020.
22	Khosla P, Teterwak P, Wang C, et al. Supervised contrastive learning[J]. Advances in Neural Information Processing Systems, 2020, 33, 18661.
23	Huang Z L, Wang X J, Huang L C, et al. CCNet: criss-cross attention for semantic segmentation[C]//IEEE/CVF International Conference on Computer Vision, 2019: 603.
24	Mao A Q, Mohri M, Zhong Y T. Cross-entropy loss functions: theoretical analysis and applications[C]//International Conference on Machine Learning, 2023: 23803.
25	Wang F, Liu H P. Understanding the behaviour of contrastive loss[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 2495.
26	Kingma D P. Adam: a method for stochastic optimization[EB/OL]. [2025-06-24]. https://doi.org/10.48550/arXiv.1412.6980.
27	John B, Menon V G, Adhikari M. FedCosA: optimised federated learning model using cosine annealing[C]//Emerging Technologies for Intelligent Systems, 2025.
28	Baldi P, Brunak S, Chauvin Y, et al. Assessing the accuracy of prediction algorithms for classification: an overview[J]. Bioinformatics, 2000, 16 (5): 412. doi: 10.1093/bioinformatics/16.5.412
29	Yacouby R, Axman D. Probabilistic extension of precision, recall, and F1 score for more thorough evaluation of classification models[C]//Workshop on Evaluation and Comparison of NLP Systems, 2020: 79.
30	Chen T, Kornblith S, Norouzi M, et al. A simple framework for contrastive learning of visual representations[C]//International Conference on Machine Learning, 2020: 1597.

参数	仿真数据集	实测数据集
载波频率	9 GHz	X波段
带宽/MHz	400	300
脉冲重复频率/Hz	800	1000
旋转角度/（°）	0.2~1.2	—

算法	准确率	F1得分
HRRPGraphNet	73.06	72.85
SimCLR	96.11	96.12
SDRnet	96.39	96.38
DPFFN	96.00	95.99
所提方法	98.06	98.07

算法	准确率	F1得分
HRRPGraphNet	73.88	64.99
SimCLR	97.76	96.51
SDRnet	98.25	96.57
DPFFN	98.02	96.65
所提方法	99.50	99.03

算法	准确率	F1得分
去掉图像编码器	76.11	76.33
去掉SSHRRP编码器	93.89	93.87
去掉MFIF模块	96.50	96.49
所提算法	98.06	98.07

算法	准确率	F1得分
去掉图像编码器	81.59	72.44
去掉SSHRRP编码器	97.76	95.41
去掉MFIF模块	98.10	97.70
所提算法	99.50	99.03

面向ISAR舰船目标的多域交互识别网络

Multi-domain interactive recognition network for ISAR ship target

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 15

参考文献 30

相关文章 15

编辑推荐

Metrics

本文评价

方法	运算时间
PCA结合随机森林	72
HRRPGraphnet	8
SimCLR	13
SDRNet	21
DPFFN	14
所提方法	11

[1]	吴少杰, 李开明, 秦钢, 张昊, 邓冬虎, 宋喜团. 基于多视角ISAR图像的雷达目标三维重建方法[J]. 系统工程与电子技术, 2026, 48(9): 2969-2978.
[2]	陈卓, 徐志明, 张杨, 艾小锋, 程江华. 空间目标极化雷达动态回波建模与仿真方法[J]. 系统工程与电子技术, 2026, 48(9): 2999-3007.
[3]	董健, 邹俊杰, 侯斐斐, 肖程望. 基于HRRP特征统计特性的雷达目标识别方法[J]. 系统工程与电子技术, 2026, 48(7): 2206-2217.
[4]	杨泽莹, 陈卓, 徐志明, 艾小锋, 吴其华, 王玲. 空间目标ISAR图像部件正向标记方法[J]. 系统工程与电子技术, 2026, 48(4): 1165-1173.
[5]	赵忠凯, 刘德雄. 基于二维相位调制的ISAR场景干扰方法[J]. 系统工程与电子技术, 2026, 48(11): 3811-3822.
[6]	蒋明煜, 张顺生, 肖思瑶. 面向轻量级交叉注意力卷积网络的SAR目标识别[J]. 系统工程与电子技术, 2025, 47(9): 2853-2861.
[7]	贾竣皓, 陈学斌, 李璋峰, 叶春茂. 空间群目标ISAR成像与检测方法[J]. 系统工程与电子技术, 2025, 47(6): 1843-1854.
[8]	蒋李兵, 杨庆伟, 郑舒予, 王壮. 基于拍卖理论的组网雷达多轨道目标ISAR成像资源分配算法[J]. 系统工程与电子技术, 2025, 47(1): 81-93.
[9]	陈洪猛, 李军, 刘京, 黄伟, 张英杰, 陈燕, 鲁耀兵. 基于Radon时频分析的海面舰船目标SAR-ISAR混合成像方法[J]. 系统工程与电子技术, 2025, 47(1): 109-116.
[10]	刘一飞, 禹卫东, 杨升辉, 李世强. 天基ISAR对空间轨道目标高分辨率成像问题分析[J]. 系统工程与电子技术, 2024, 46(11): 3671-3683.
[11]	潘小义, 王泉, 顾赵宇, 陈吉源, 孟晓明. 基于调频斜率变化的ISAR成像干扰方法研究[J]. 系统工程与电子技术, 2024, 46(11): 3703-3714.
[12]	寇鹏, 刘永祥, 张弛, 李玮杰, 张双辉, 霍凯. 序列ISAR像复杂结构航天器在轨姿态估计[J]. 系统工程与电子技术, 2023, 45(8): 2438-2445.
[13]	朱瀚神, 胡文华, 郭宝锋, 焦丽婷, 朱晓秀, 朱常安. 双基地ISAR稀疏孔径机动目标MTRC补偿成像算法[J]. 系统工程与电子技术, 2023, 45(7): 2022-2030.
[14]	杨磊, 夏亚波, 廖仙华, 毛欣瑶, 窦宇宸, 杨桓. 双层稀疏贝叶斯学习ISAR超分辨成像算法[J]. 系统工程与电子技术, 2023, 45(5): 1371-1379.
[15]	吴其华, 赵锋, 王俊杰, 徐志明, 肖顺平. 基于编码相位调制的ISAR图像特征控制方法[J]. 系统工程与电子技术, 2023, 45(4): 1000-1007.