| 22 | 0 | 57 |
| 下载次数 | 被引频次 | 阅读次数 |
水下声呐目标检测在推动海洋资源勘探、促进环境监测，以及保障水下作业安全等方面具有不可替代的重要作用。随着技术的发展，目标检测方法已从依赖人工设计特征转向基于深度学习的自动特征提取与识别。然而，声呐图像的低分辨率与高噪声特性，使得深度模型在提取与融合有效特征方面依然面临挑战。为此，本文提出一种基于自调节特征增强和迭代注意力的水下声呐目标检测方法。该方法提出自调节特征增强模块，用于提升对关键信息的感知能力；设计了多尺度空间特征精炼模块，实现对多尺度目标的有效检测；同时提出迭代式多维注意力融合模块，实现深浅层特征的高效融合，从而增强模型在复杂水下场景下的检测准确率。实验结果表明，在3个不同的开源数据集上，本文方法的mAP@75指标分别为61.4%,34.9%和61.8%,均优于对比方法。
Abstract：Underwater sonar target detection plays an irreplaceable role in advancing marine resource exploration, facilitating environmental monitoring, and ensuring the safety of underwater operations. With the rapid development of technology, detection methods have evolved from relying on handcrafted features to deep learning-based approaches capable of automatic feature extraction and recognition. However, the low resolution and high noise of sonar imagery still pose significant challenges for deep models in effectively extracting and fusing discriminative features. To overcome these limitations, this paper proposes an underwater sonar target detection method based on self-regulating feature enhancement and iterative attention. Specifically, a self-regulating feature enhancement module is designed to strengthen the perception of key information, a multi-scale spatial feature refinement module is introduced to achieve robust detection of targets at multiple scales, and an iterative multi-dimensional attention fusion module is developed to efficiently integrate deep and shallow features, thereby improving detection accuracy in complex underwater environments. Experimental results show that on three different open-source datasets, our method achieves mAP@75 scores of 61.4%, 34.9%, and 61.8%, all of which surpass comparative methods.
[1] Zou Z,Shi Z,Guo Y,et al.Object detection in 20 years:A survey[C]//Proceedings of the IEEE.Piscataway:IEEE,2023,111(3):257-276.
[2] Mallat S,Zhong S.Characterization of signals from multiscale edges[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1992,14(7):710-732.
[3] LeCun Y,Bengio Y,Hinton G.Deep learning[J].Nature,2015,521(7553):436-444.
[4] Alzubaidi L,Zhang J,Humaidi A J,et al.Review of deep learning:Concepts,CNN architectures,challenges,applications,future directions[J].Journal of Big Data,2021,8(1):1-74.
[5] Krizhevsky A,Sutskever I,Hinton G E.ImageNet classification with deep convolutional neural networks[J].Communications of the ACM,2017,60(6):84-90.
[6] He K,Gkioxari G,Dollár P,et al.Mask R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision.Piscataway:IEEE,2017:2961-2969.
[7] Fang Y,Sun Y,Wang Q,et al.You only look at one sequence:Rethinking transformer in vision through object detection[C]//Advances in Neural Information Processing Systems.Red Hook:Curran Associates,2021,34:26183-26197.
[8] Ma Q,Jiang L,Yu W,et al.Training with noise adversarial network:A generalization method for object detection on sonar image[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.Piscataway:IEEE,2020:729-738.
[9] Aubard M,Antal L,Madureira A,et al.ROSAR:An adversarial re-training framework for robust side-scan sonar object detection[EB/OL].arXiv preprint arXiv:2410.10554.(2024-10-15) [2025-10-11].https://arxiv.org/abs/2410.10554.
[10] Tang Y,Wang L,Li H,et al.Side-scan sonar underwater target segmentation using the BHP-UNet[J].EURASIP Journal on Advances in Signal Processing,2023,2023(1):76.
[11] Wang Z,Zhang S.Sonar image detection based on multi-scale multi-column convolution neural networks[J].IEEE Access,2019,7:160755-160767.
[12] Shi P,Sun H,Xin Y,et al.SDNet:Image-based sonar detection network for multi-scale objects[J].IET Image Processing,2023,17(4):1208-1223.
[13] Zou C,Yu S,Yu Y,et al.Side-scan sonar small objects detection based on improved YOLOv11[J].Journal of Marine Science and Engineering,2025,13(1):162.
[14] Yu F,Koltun V.Multi-scale context aggregation by dilated convolutions[EB/OL].arXiv preprint arXiv:1511.07122.(2015-11-23) [2025-10-11].https://arxiv.org/abs/1511.07122.
[15] Zhu X,Su W,Lu L,et al.Deformable DETR:Deformable transformers for end-to-end object detection[EB/OL].arXiv preprint arXiv:2010.04159.(2020-10-08) [2025-10-11].https://arxiv.org/abs/2010.04159.
[16] Hu J,Shen L,Sun G.Squeeze-and-excitation networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE,2018:7132-7141.
[17] Khanam R,Hussain M.YOLOv11:An overview of the key architectural enhancements[EB/OL].arXiv preprint arXiv:2410.17725.(2024-10-28) [2025-10-11].https://arxiv.org/abs/2410.17725.
[18] Dai Y,Gieseke F,Oehmcke S,et al.Attentional feature fusion[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.Piscataway:IEEE,2021:3560-3569.
[19] Wu Y,He K.Group normalization[C]//Proceedings of the European Conference on Computer Vision.Cham:Springer,2018:3-19.
[20] 周彦，陈少昌，吴可，等.SCTD1.0:声呐常见目标检测数据集[J].计算机科学，2021,48(S2):334-339.
[21] University of Jinan.dataset1 Dataset[EB/OL].(2024-09)/[2025-04-22].https://universe.roboflow.com/university-of-jinan/dataset1-ixoyo.
[22] IIT Palakkad.Sonar Detection Dataset[EB/OL].(2023-04-01)/[2025-10-11].https://universe.roboflow.com/iit-palakkad/sonar-detection.
[23] Selvaraju R R,Cogswell M,Das A,et al.Grad-CAM:Visual explanations from deep networks via gradient-based localization[C]//Proceedings of the IEEE International Conference on Computer Vision.Piscataway:IEEE,2017:618-626.
[24] Wang Z,Guo J,Zhang S,et al.Marine object detection in forward-looking sonar images via semantic-spatial feature enhancement[J].Frontiers in Marine Science,2025,12:1598765.
[25] Xing B,Sun M,Ding M,et al.Fish sonar image recognition algorithm based on improved YOLOv5[J].Mathematical Biosciences and Engineering,2024,21(1):234-250.
[26] Li X,Zhao Y,Su H,et al.Efficient underwater object detection based on feature enhancement and attention detection head[J].Scientific Reports,2025,15(1):1-15.
(1)本方法代码发布在:https://github.com/miraclebai/YOLOv11-MFA。
基本信息:
DOI：10.20064/j.cnki.2095-347X.2026.02.003
中图分类号:TP391.41;TB566
引用信息:
[1]白京源 ,尹臣杰 ,王津港 ,等.基于自调节特征增强与迭代注意力的声呐目标检测方法[J].网络新媒体技术,2026,15(02):19-29.DOI:10.20064/j.cnki.2095-347X.2026.02.003.
基金信息:
海南省科技专项(编号：ZDYF2025SHFZ058); 中国科学院青年创新促进会项目(编号：2022022); 海南省“南海新星”科技创新人才平台项目(编号：NHXXRCXM202340); 海口市重点科技计划项目(编号：2024020)
2026-03-15
2026-03-15