针对复杂水域场景中船舶区域分割研究不足、目标特征弱化及尺寸差异性显著导致的算法分割精度低、计算效率差等问题，本文提出基于YOLO11n-seg框架改进的YOLO-OACE分割网络。首先，设计全维动态特征提取模块以增强骨干网络对船舶目标多维度特征的捕捉能力；其次，引入轻量化ADown下采样策略，通过降维压缩特征图尺寸，在减少模型参数量与计算复杂度的同时保留关键信息表征；进一步结合组注意力机制，强化网络对多尺度船舶区域的动态聚焦与分割适应性；最后，采用EIOU损失函数优化边界框回归机制，通过改进宽高约束项提升分割精度与收敛效率。实验表明，相较于YOLO11n-seg模型，YOLO-OACE平均分割精度提升4.7%，计算量降低21%，参数量减少16%，模型体积缩减0.8 MB，显著优于现有方法，可为复杂背景下船舶智能化识别和水面视觉导航提供技术支撑。
To address the challenges of low segmentation accuracy and high computational cost caused by insufficient research on ship area segmentation, weakened feature representations, and large-scale variability in complex water scenes, this paper proposes an improved segmentation network named YOLO-OACE, based on the YOLO11n-seg framework. First, an omni-dimensional dynamic feature extraction module is introduced to enhance the backbone’s capacity for capturing multi-scale and multi-dimensional features of ship targets. Second, a lightweight ADown downsampling strategy compresses the feature maps while retaining critical information, effectively reducing model parameters and computational complexity. Furthermore, a group attention mechanism is integrated to improve the network’s ability to adaptively focus on ship regions across different scales. Finally, the EIoU loss function is adopted to optimize bounding box regression by strengthening width and height constraints, thereby enhancing segmentation precision and convergence speed. Experimental results show that, compared to the original YOLO11n-seg model, the proposed YOLO-OACE achieves a 4.7% improvement in average segmentation accuracy, reduces computation by 21%, decreases parameters by 16%, and shrinks model size by 0.8 MB. These results demonstrate that YOLO-OACE significantly outperforms existing methods and offers strong potential for intelligent ship detection and visual navigation in complex maritime environments.
2026,48(2): 138-144 收稿日期：2025-5-14
DOI：10.3404/j.issn.1672-7649.2026.02.022
分类号：U662.9；TP391.4
基金项目：国家自然科学基金资助项目（52001235）；湖北省自然科学基金资助项目（2022CFB313）
作者简介：陈赫翔(2001-),男,硕士研究生,研究方向为多模态船舶检测与定位、计算机视觉
参考文献：
[1] 黄琛, 陈德山, 吴兵, 等. 船舶航行交通事件实时检测技术研究现状与展望[J]. 交通信息与安全, 2022, 40(6): 1-11.
HUANG Chen, CHEN Deshan, WU Bing, et al. A real-time detection of nautical traffic events: A review and prospect[J]. Journal of Transport Information and Safety, 2022, 40(6): 1-11.
[2] HORDIIUK D, OLIINYK I, HNATUSHENKO V, et al. Semantic segmentation for ships detection from satellite imagery[C]//2019 IEEE 39th International Conference on Electronics and Nanotechnology (ELNANO). IEEE, 2019: 454-457.
[3] CHATURVEDI S K. Study of synthetic aperture radar and automatic identification system for ship target detection[J]. Journal of Ocean Engineering and Science, 2019, 4(2): 173-182.
[4] GAO Z, LIU Z, ZHANG T. Infrared ship target segmentation based on adversarial domain adaptation[J]. Data Acquis. Process., 2023, 38(3): 598-607.
[5] DAI J, HE K, LI Y, et al. Instance-sensitive fully convolutional networks[C] //Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part VI 14. Springer International Publishing, 2016: 534-549.
[6] HARIHARAN B, ARBELÁEZ P, GIRSHICK R, et al. Simultaneous detection and segmentation[C] //Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VII 13. Springer International Publishing, 2014: 297-312.
[7] REN S, HE K, GIRSHICK R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[J]. IEEE transactions on pattern analysis and machine intelligence, 2016, 39(6): 1137-1149.
[8] HE K, GKIOXARI G, DOLLÁR P, et al. Mask r-cnn[C]//Proceedings of the IEEE international conference on computer vision. 2017: 2961-2969.
[9] 张文凯, 余敏, 刘浩煜, 等. 低检测置信度下轻量化水下多目标跟踪算法[J]. 舰船科学技术, 2025, 47(6): 128-133.
ZHANG W K, YU M, LIU H Y, et al. A light underwater multi-object tracking algorithm with low detection confidence[J]. Ship Science and Technology, 2025, 47(6): 128-133.
[10] WANG X, KONG T, SHEN C, et al. Solo: Segmenting objects by locations[C]//Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XVIII 16. Springer International Publishing, 2020: 649-665.
[11] WANG X, ZHANG R, KONG T, et al. Solov2: Dynamic and fast instance segmentation[J]. Advances in Neural information processing systems, 2020, 33: 17721-17732.
[12] BOLYA D, ZHOU C, XIAO F, et al. Yolact: Real-time instance segmentation[C]//Proceedings of the IEEE/CVF international conference on computer vision, 2019: 9157-9166.
[13] HUANG M, XU G, LI J, et al. A method for segmenting disease lesions of maize leaves in real time using attention YOLACT++[J]. Agriculture, 2021, 11(12): 1216.
[14] KE X, ZHANG T, SHAO Z. Scale-aware dimension-wise attention network for small ship instance segmentation in synthetic aperture radar images[J].Journal of Applied Remote Sensing, 2023, 17(4): 046504-046504.
[15] TAN S, YAN J, JIANG Z, et al. Approach for improving YOLOv5 network with application to remote sensing target detection[J]. Journal of Applied Remote Sensing, 2021, 15(3): 036512-036512.
[16] 孙雨鑫, 苏丽, 陈禹升, 等. 基于注意力机制的 SOLOA 船舶实例分割算法[J]. 智能系统学报, 2023, 18(6): 1197-1204.
SUN Y X, SU L, CHEN Y S, et al. SOLOA ship instance segmentation algorithm based on attention[J]. CAAI transactions on intelligent systems, 2023, 18(6): 1197-1204.
[17] 王磊, 张斌, 吴奇鸿. RCSA-YOLO: 改进 YOLOv8 的 SAR 舰船实例分割[J]. 计算机工程与应用, 2024, 60(18): 1.3–113.
WANG L, ZHANG B, WU Q H. RCSA-YOLO: Improved SAR Ship Instance Segmentation of YOLOv8[J]. Journal of Computer Engineering & Applications, 2024, 60(18): 1.3–113.
[18] HUANG Y, WANG D, WU B, et al. NST-YOLO11: vit merged model with neuron attention for arbitrary-oriented ship detection in SAR images[J]. Remote Sensing, 2024, 16(24): 4760.
[19] GAN Y, REN X, LIU H, et al. A novel lightweight YOLO11-based framework for precisely locating diverse ship targets in complex optical remote sensing photographs[J]. Measurement Science and Technology, 2025, 36(4): 045409.
[20] LI C, ZHOU A, YAO A. Omni-dimensional dynamic convolution[J]. arXiv preprint arXiv: 2209.07947, 2022.
[21] WANG C Y, YEH I H, MARK LIAO H Y. Yolov9: Learning what you want to learn using programmable gradient information[C]//European conference on computer vision. Cham: Springer Nature Switzerland, 2024: 1-21.
[22] LIU X, PENG H, ZHENG N, et al. Efficientvit: Memory efficient vision transformer with cascaded group attention[C] //Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2023: 14420-14430.
[23] ZHENG Z, WANG P, REN D, et al. Enhancing geometric factors in model learning and inference for object detection and instance segmentation[J]. IEEE transactions on cybernetics, 2021, 52(8): 8574-8586.
[24] ZHANG Y F, REN W, ZHANG Z, et al. Focal and efficient IOU loss for accurate bounding box regression[J]. Neurocomputing, 2022, 506: 146-157.
[25] SHAO Z, WU W, WANG Z, et al. Seaships: a large-scale precisely annotated dataset for ship detection[J]. IEEE transactions on multimedia, 2018, 20(10): 2593-2604.
