航空学报 > 2026, Vol. 47 Issue (17): 333230-333230   doi: 10.7527/S1000-6893.2026.33230

基于跨域频空先验迁移的SAR图像分割

夏梓涵1, 刘瑾2(), 黄星儒1, 郑志文1, 张晓帅3   

  1. 1.杭州电子科技大学 通信工程学院,杭州 310018
    2.武汉大学 遥感信息工程学院,武汉 430079
    3.中国海洋大学 信息科学与工程学院,青岛 266404
  • 收稿日期:2025-12-15 修回日期:2025-12-31 接受日期:2026-03-04 出版日期:2026-03-20 发布日期:2026-03-19
  • 通讯作者: 刘瑾 E-mail:jinliu@hdu.edu.cn
  • 基金资助:
    国家自然科学基金青年基金(42501545);国家自然科学基金青年基金(62506103);中国博士后科学基金(2024M760716);中国博士后科学基金特别资助(2025T180962);浙江省属高校基本科研业务费(GK259909299001-026);山东省海外优秀青年科学基金(2025HWYQ-033)

SAR image segmentation based on cross-domain transfer of frequency-spatial priors

Zihan XIA1, Jin LIU2(), Xingru HUANG1, Zhiwen ZHENG1, Xiaoshuai ZHANG3   

  1. 1.School of Communication Engineering,Hangzhou Dianzi University,Hangzhou 310018,China
    2.Remote Sensing and Information Engineering,Wuhan University,Wuhan 430079,China
    3.Faculty of Information Science and Engineering,Ocean University of China,Qingdao 266404,China
  • Received:2025-12-15 Revised:2025-12-31 Accepted:2026-03-04 Online:2026-03-20 Published:2026-03-19
  • Contact: Jin LIU E-mail:jinliu@hdu.edu.cn
  • Supported by:
    National Natural Science Foundation of China(42501545);China Postdoctoral Science Foundation(2024M760716);Special Financial Grant from the China Postdoctoral Science Foundation(2025T180962);Fundamental Research Funds for the Provincial Universities of Zhejiang(GK259909299001-026);Excellent Young Scientists Fund Program (Overseas) of Shandong Province(2025HWYQ-033)

摘要:

合成孔径雷达(SAR)凭借全天候成像优势广泛应用于对地观测,但其成像机制固有的斑点噪声、强背景杂波干扰以及缺乏直观几何纹理的特性,导致复杂散射场景下的高精度语义分割面临巨大挑战。现有深度学习方法多依赖各向同性卷积,难以有效解耦频域噪声与空间结构,基于光学先验引入的跨域辅助分割方案往往受限于对严格配准光学-SAR成对数据的依赖,限制了其适用性。为此,提出一种语义归纳双域各向重构GAN框架(SIDAR-GAN)。该方法突破了对严格配准数据的依赖,设计了基于生成对抗网络(GAN)的跨域一致性学习机制。通过构建语义归纳GAN(SI-GAN)与双域结构对齐(DDSA)模块,从生成器编码层提取隐含的光学几何与纹理先验,并通过跨域注意力机制将其自适应注入SAR特征空间,实现了非配准条件下的高鲁棒性语义归纳。进一步设计的频-空扰度解析模块(FSDP),通过频域扰度解析分支利用小波多尺度分解实现噪声的频域隔离与自适应抑制,同时利用空间各向解析分支引入形状偏置卷积核,强化目标(如舰船、建筑)在惯性方向上的特征表达,从频空双域联合建模结构与噪声特性。在SARBuD建筑数据集与HRSID舰船数据集上的实验表明,该方法Dice系数分别达到77.66%和92.28%,与18种主流模型相比,噪声抑制能力与边界提取精度显著提升,为弱纹理条件下的SAR精细化解译提供了新的跨模态语义驱动范式。

关键词: 合成孔径雷达, 图像分割, 生成对抗网络, 噪声抑制, 深度学习

Abstract:

Synthetic Aperture Radar (SAR) is widely used in Earth observation due to its all-weather imaging capability. However, its inherent speckle noise, strong background clutter, and lack of intuitive geometric texture make high-precision semantic segmentation in complex scattering scenes highly challenging. Existing deep learning methods mostly rely on isotropic convolutions, which are insufficient for effectively disentangling frequency-domain noise from spatial structures. Cross-domain auxiliary segmentation schemes based on optical priors are often limited by their dependence on strictly aligned optical-SAR paired data, which restricts their applicability. To address these issues, this paper proposes a Semantic Induction Dual-domain Anisotropic Reconstruction with Generative Adversarial Networks (SIDAR-GAN). The proposed method breaks the dependence on strictly aligned data and designs a cross-domain consistency learning mechanism based on a Generative Adversarial Network (GAN). By constructing a Semantic Induction GAN (SI-GAN) and a Dual-Domain Structural Alignment (DDSA) module, implicit optical geometric and texture priors are extracted from the generator encoder and adaptively injected into the SAR feature space through a cross-domain attention mechanism, enabling robust semantic induction under unregistered conditions. Furthermore, a Frequency-Spatial Disturbance Parsing (FSDP) module is designed. Its frequency-domain branch employs multi-scale wavelet decomposition to isolate and adaptively suppress noise, while its spatial anisotropic branch introduces shape-biased convolution kernels to enhance the feature representation of targets such as ships and buildings along their principal directions, thereby jointly modeling structural and noise characteristics in both frequency and spatial domains. Experiments on the SARBuD building dataset and the HRSID ship dataset demonstrate that the proposed method achieves Dice coefficients of 77.66% and 92.28%, respectively. Compared with 18 mainstream models, it significantly improves noise suppression capability and boundary extraction accuracy, providing a new cross-modal semantic-driven paradigm for fine-grained SAR interpretation under weak-texture conditions.

Key words: synthetic aperture radar, image segmentation, generative adversarial network, noise suppression, deep learning

中图分类号: