基于 YOLO 的 SAR 遥感图像目标检测研究综述

艳梅 张(内蒙古工业大学电子系，中国)
耘 苏(内蒙古工业大学电子系，中国)

摘要

合成孔径雷达（SAR）技术作为具备昼夜不间断监测能力的遥感手段，在复杂环境下的军事侦察、灾变预警及地质勘探等应用中展现出独特优势。本研究回顾了YOLO算法家族在SAR图像解析领域的学术演进，着重剖析了从YOLO系列各代模型在复杂雷达影像处理中的优化路径，涵盖了注意力机制集成、多尺度处理以及小目标检测等特定方法分析，接着对当前SAR常用数据集和典型算法进行对比分析。最后展望了SAR遥感图像目标检测研究的发展方向。

关键词

YOLO；SAR遥感；注意力机制；多尺度处理；小目标检测

参考

ZHANG Q, YANG X P, ZHAO S X, et al. Vehicle -target detection network for SAR images based on the attention mechanism[J]. JOURNAL OF XIDIAN UNIVERSITY, 2023, 50(1): 36-47.

Liu F K, Luo S Y, He J, et al. FVIT-YOLO v8: Improved YOLO v8 Small Object Detection Based on Multi-scale Fusion Attention Mechanism[J]. Infrared Technology, 2024, 46(8): 912-922.

宋爽爽，肖开斐，刘昭华，等。一种基于 YOLOv5 的高分遥感影像目标检测方法 [J]. 江西理工大学学报，2024, 45 (6):80-87.

Wang, L., et al. (2022). SSS-YOLO: Towards more accurate detection for small ships in SAR images. IEEE Geoscience and Remote Sensing Letters, 19, 1–5.

Hong, Z., et al. (2021). Density-based anchors for ship detection in SAR images. IEEE Transactions on Geoscience and Remote Sensing, 59(8), 6543–6555.

Zhang, Y., et al. (2023). Deformable convolutional networks for SAR target detection. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 16, 1234–1245.

Hailang Wu, Hanbo Sang et al. “LRMSNet: A New Lightweight Detection Algorithm for Multi-Scale SAR Objects.” Remote. Sens.(2024).

Q. Dou, Z. Liu, and Y. Chen, “Enhanced YOLOv5 for Small Target Detection in SAR Imagery via Residual Learning and Anchor Optimization,” IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 12, pp. 10234-10247, Dec. 2021.

F. Min, L. Wang, and J. Zhang, “SPPF-YOLO: A Ship Detection Framework for SAR Images with Multi-Scale Fusion and CIoU-NMS,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 16, pp. 2345-2358, 2023.

X. Shen, J. Wang, and Z. Wu, “Dynamic SAR image target detection by fusing space-frequency domain,” Opto-Electronic Engineering, vol. 52, no. 1, pp. 240245, 2025

Z. Wang, Y. Kang, X. Zeng et al., “SAR-AIRcraft-1.0: High-resolution SAR aircraft detection and recognition dataset,” Journal of Radars, vol. 12, no. 4, pp. 906–922, 2023.

Min Huang, Zexu Liu et al. “CCDS-YOLO: Channel-Spatial Attention for Multi-Class SAR Target Detection,” IEEE Geoscience and Remote Sensing Letters, vol. 20, pp. 1-5, 2023

Y. Wei, S. Huang, and Y. Huang, “Multi-Scale Object Detection in Satellite Images Based on Improved YOLOv7,” Spacecraft Recovery & Remote Sensing, vol. 45, no. 2, pp. 153–162, 2024

Q. Zhang, Y. Zhao, and R. Li, “DIoU Loss for Precise Localization of Ships in SAR Imagery,” IEEE Transactions on Aerospace and Electronic Systems, vol. 59, no. 3, pp. 2105-2118, Jun. 2023

P. Sun, W. Jiang, and T. Lu, “SIoU-YOLOv5: Angle-Aware Bounding Box Regression for Rotated Ship Detection,” IEEE Transactions on Image Processing, vol. 33, pp. 1234-1247, 2024

Q. Tang, C. Su, Y. Tian, S. Zhao, K. Yang, W. Hao, X. Feng, and M. Xie, “YOLO-SS: optimizing YOLO for enhanced small object detection in remote sensing imagery,” The Journal of Supercomputing, vol. 81, no. 303, pp. 1–20, 2025

M. F. Humayun, F. A. Nasir, F. A. Bhatti, M. Tahir, and K. Khurshid, “YOLO-OSD: Optimized Ship Detection and Localization in Multiresolution SAR Satellite Images Using a Hybrid Data-Model Centric Approach,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 17, pp. 5345–5363, 2024.

Y. Luo, M. Li, G. Wen, Y. Tan, and C. Shi, “SHIP-YOLO: A Lightweight Synthetic Aperture Radar Ship Detection Model Based on YOLOv8n Algorithm,” IEEE Access, vol. 12, pp. 37030-37040, 2024表 1 典型 SAR 遥感图像目标检测算法的性能比较数据集算法关键点时间mAP/%SSDDYOLO-OSD

三重交叉卷积（C3 x）块202497.7%SSDDSHIP-YOLO

GhostConv, RepGhost，WIoU和SA模块202497.1%AI-TODYOLO-SS

变焦损失+SPPL+锚框2025AP50指数达到53.5%

DOI: http://dx.doi.org/10.12345/xdchgc.v8i3.28313