Keywords
Unsupervised domain adaptation
SAR
Marine aquaculture
Submitted : 2026-06-23
Accepted : 2026-07-08
Published : 2026-07-23
Abstract
Marine aquaculture is an important part of the marine economy in China. Synthetic aperture radar (SAR) images can work at all times, which makes them suitable for large-scale monitoring of aquaculture areas. However, SAR images from different regions often have obvious differences in background scattering, aquaculture structures, and imaging conditions. These differences cause serious domain shift problems and reduce the segmentation performance of existing methods in cross-region tasks. To solve this problem, this paper proposes a structure- and frequency-aware domain adaptation method (SFDA) for raft aquaculture extraction from SAR images. The method includes a directional structure alignment (DSA) module and a frequency domain alignment (FDA) module. The DSA module aligns the directional structure features between the source and target domains, while the FDA module aligns texture and scattering features in the frequency domain. Experiments on the Gaofen-3 SAR dataset from Jiangsu province show that the proposed method is effective for cross-region raft aquaculture extraction.
References
Pörtner H O, Scholes R, Arneth A, et al., 2023, Overcoming the Coupled Climate and Biodiversity Crises and Their Societal Impacts. Science, 380(6642): eabl4881.
Fan J, Zhao J, Song D, et al., 2018, Marine Floating Raft Aquaculture Dynamic Monitoring Based on Multi-Source GF Imagery. 2018 7th International Conference on Agro-Geoinformatics, 1–4.
Chen W, Li X, 2024, Deep-Learning-Based Marine Aquaculture Zone Extractions from Dual-Polarimetric SAR Imagery. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 17: 8043–8057.
Wang J, Fan J, Wang J, 2022, MDOAU-Net: A Lightweight and Robust Deep Learning Model for SAR Image Segmentation in Aquaculture Raft Monitoring. IEEE Geoscience and Remote Sensing Letters, 19: 1–5.
Fan J, Zhao J, An W, et al., 2019, Marine Floating Raft Aquaculture Detection of GF-3 PolSAR Images Based on Collective Multikernel Fuzzy Clustering. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 12(8): 2741–2754.
Wang X, Zhou J, Fan J, 2022, IDUDL: Incremental Double Unsupervised Deep Learning Model for Marine Aquaculture SAR Images Segmentation. IEEE Transactions on Geoscience and Remote Sensing, 60: 1–12.
Zhao J, Li Y, Zhou Y, et al., 2025, DDCI: Unsupervised Domain Adaptation for Remote Sensing Images Based on Diffusion Causal Distillation. IEEE Transactions on Geoscience and Remote Sensing, 63: 1–12.
Zhu J, Guo Y, Sun G, et al., 2024, Causal Prototype Inspired Contrast Adaptation for Unsupervised Domain Adaptive Semantic Segmentation of High-Resolution Remote Sensing Imagery. IEEE Transactions on Geoscience and Remote Sensing, 62: 1–17.
Zhu C, Liu K, Tang W, et al., 2025, Hard-Aware Instance Adaptive Self-Training for Unsupervised Cross-Domain Semantic Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 47(7): 5655–5671.
Shi Y, Du L, Li C, et al., 2024, Unsupervised Domain Adaptation for SAR Target Classification Based on Domain- and Class-Level Alignment: From Simulated to Real Data. ISPRS Journal of Photogrammetry and Remote Sensing, 207: 1–13.
Yang Y, Chen J, Sun L, et al., 2024, Unsupervised Domain-Adaptive SAR Ship Detection Based on Cross-Domain Feature Interaction and Data Contribution Balance. Remote Sensing, 16(2): 420.
Liu S, Li D, Song H, et al., 2025, SAR Ship Detection Across Different Spaceborne Platforms with Confusion-Corrected Self-Training and Region-Aware Alignment Framework. ISPRS Journal of Photogrammetry and Remote Sensing, 228: 305–322.
Yang Y, Yang X, Yang D, 2025, Unsupervised Domain Adaptation for SAR Ship Detection Based on Multitask Decoupling. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 18: 12684–12696.
Tu H, Wang W, Guo Y, et al., 2025, Mamba-UDA: Mamba Unsupervised Domain Adaptation for SAR Ship Detection. IEEE Geoscience and Remote Sensing Letters, 22: 1–5.
Ren Z, Du Z, Zhang Y, et al., 2024, Multi-Step Unsupervised Domain Adaptation in Image and Feature Space for Synthetic Aperture Radar Image Terrain Classification. Remote Sensing, 16(11): 1901.
Cui G, Fan J, Zou Y, 2025, Enhanced Unsupervised Domain Adaptation with Iterative Pseudo-Label Refinement for Inter-Event Oil Spill Segmentation in SAR Images. International Journal of Applied Earth Observation and Geoinformation, 139: 104479.
Vu T, Jain H, Bucher M, et al., 2019, ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2512–2521.
Arthur G, Karsten M B, Malte J R, et al., 2012, A Kernel Two-Sample Test. Journal of Machine Learning Research, 13(25): 723–773.
Fan J, Li M, Wang X, 2025, Unsupervised Transformer with Generative Label Optimization for Marine Aquaculture Segmentation. IEEE Transactions on Geoscience and Remote Sensing, 63: 1–14.
Zhang L, Lan M, Zhang J, et al., 2022, Stagewise Unsupervised Domain Adaptation with Adversarial Self-Training for Road Segmentation of Remote-Sensing Images. IEEE Transactions on Geoscience and Remote Sensing, 60: 1–13.