[1]沈昊,葛泉波,吴高峰.基于主干网络浅深层特征的无人机海上分割算法[J].智能系统学报,2025,20(3):605-620.[doi:10.11992/tis.202405021]
SHEN Hao,GE Quanbo,WU Gaofeng.Unmanned aerial vehicle-driven sea segmentation based on the shallow and deep features of the backbone[J].CAAI Transactions on Intelligent Systems,2025,20(3):605-620.[doi:10.11992/tis.202405021]
点击复制
SHEN Hao,GE Quanbo,WU Gaofeng.Unmanned aerial vehicle-driven sea segmentation based on the shallow and deep features of the backbone[J].CAAI Transactions on Intelligent Systems,2025,20(3):605-620.[doi:10.11992/tis.202405021]
基于主干网络浅深层特征的无人机海上分割算法
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
20
期数:
2025年第3期
页码:
605-620
栏目:
学术论文—机器学习
出版日期:
2025-05-05
- Title:
- Unmanned aerial vehicle-driven sea segmentation based on the shallow and deep features of the backbone
- Keywords:
- complex maritime scene; semantic segmentation; drone landing; ship target; DeepLabV3+; attention mechanism; deep learning; convolutional neural network
- 分类号:
- TP751
- 摘要:
- 为提高复杂海洋环境中无人机自主降落时分割目标的实时性和精确性，研究主干网络和浅深层特征对分割算法性能的影响问题，基于DeepLabV3+框架建立一种基于主干网络浅深层特征的无人机海上分割(shallow and deep features of backbone, SDFB)算法。首先，针对风浪扰动降低目标稳定性的问题，优化MobileNetV2结构提出一种特征提取方法，解决了算法无法处理短时间目标变化较大图像的问题；然后，针对深层特征输出通道数较多且存在不均匀分布大气湍流噪声的问题，利用本地全局信息选择性地聚合特征，提出一种特征筛选机制，剔除冗余通道的同时解决了算法对环境噪声敏感度高的问题；其次，针对光照不匀降低目标边界清晰度问题，从浅层空间维度和深层通道维度中提取轮廓信息建立一种并行轮廓学习机制，解决了算法利用轮廓特征效率低的问题；最后，针对障碍物遮挡破坏目标特征完整性问题，融合优化后的条带池化建立一种特征融合机制，解决了算法无法联系离散分布特征问题。实验表明，SDFB算法的实时性和精确性均高于其他算法，能够更好地适应海上场景无人机分割目标需求。
- Abstract:
- To improve the real-time and accurate segmentation of targets during the autonomous landing of UAVs in complex marine environments, studying the impact of backbone and shallow/deep features on the performance of algorithms is crucial. Based on the DeepLabV3+ framework, a Shallow and Deep Features of Backbone (SDFB) algorithm is established for maritime scene segmentation. First, to address the issue of reduced target stability caused by wind-wave disturbances, a feature extraction method is proposed by optimizing the MobileNetV2 structure, and this method resolves the issue of low processing speed of single frame images in the algorithm. Second, to address the issue of numerous deep feature output channels and the uneven distribution of atmospheric turbulence noise, a feature filtering mechanism is proposed by selectively aggregating features using local and global information, thereby eliminating redundant features while solving the high sensitivity issue of the algorithm to environmental noise. Third, to address the issue of uneven lighting reducing the clarity of target boundaries, a parallel contour learning mechanism is established by extracting contour information from shallow spatial dimensions and deep channel dimensions, thereby solving the low-efficiency issue regarding the utilization of contour features. Finally, to address the issue of background occlusion disrupting the integrity of target features, a multi-scale feature fusion mechanism is established through the fusion optimization of strip pooling, and this solves the connection issue of the algorithm to discrete distribution features. Finally, relevant experiments reveal that the LMSC algorithm exhibits higher real-time accuracy than other algorithms and can better adapt to the segmentation requirements of UAVs in maritime scenes.
- 参考文献/References:
-
[1] LIU Yalin, DAI Hongning, WANG Q, et al. Unmanned aerial vehicle for Internet of everything: opportunities and challenges[J]. Computer communications, 2020, 155: 66-83.
[2] RAOULT V, COLEFAX A P, ALLAN B M, et al. Operational protocols for the use of drones in marine animal research[J]. Drones, 2020, 4(4): 64.
[3] 饶颖露, 邢金昊, 张恒, 等. 基于视觉的无人机板载自主实时精确着陆系统[J]. 计算机工程, 2021, 47(10): 290-297.
RAO Yinglu, XING Jinhao, ZHANG Heng, et al. Vision-based autonomous real-time precise landing system for UAV-borne processors[J]. Computer engineering, 2021, 47(10): 290-297.
[4] 原辉, 裴楚, 王帅, 等. 基于地标的轻量化无人机精准降落算法[J]. 计算机工程与应用, 2022, 58(17): 306-313.
YUAN Hui, PEI Chu, WANG Shuai, et al. Landmark-based lightweight UAV accurate landing algorithm[J]. Computer engineering and applications, 2022, 58(17): 306-313.
[5] 金建海, 周则兴, 张波, 等. 无人艇航行仿真关键技术研究[J]. 系统仿真学报, 2021, 33(12): 2846-2853.
JIN Jianhai, ZHOU Zexing, ZHANG Bo, et al. Research on USV navigation simulation key technologies[J]. Journal of system simulation, 2021, 33(12): 2846-2853.
[6] MAO Zhiyuan, CHIMITT N, CHAN S H. Accelerating atmospheric turbulence simulation via learned phase-to-space transform[C]//2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 14739-14748.
[7] KUBRYAKOV A A, KUDRYAVTSEV V N, STANICHNY S V. Application of Landsat imagery for the investigation of wave breaking[J]. Remote sensing of environment, 2021, 253: 112144.
[8] GHAZALI M H M, RAHIMAN W. An investigation of the reliability of different types of sensors in the real-time vibration-based anomaly inspection in drone[J]. Sensors, 2022, 22(16): 6015.
[9] 龙翔, 陈华杰, 吴浩宇, 等. 基于特征增强的海面强干扰目标检测[J]. 激光与光电子学进展, 2025, 62(2): 1-20.
Long Xiang, Chen Huajie, Wu Haoyu, et al. Strong Interference Target Detection on the Sea Surface Based on Feature Augmentation[J]. Laser & optoelectronics progress, 2025, 62(2): 1-20.
[10] 李伟, 杨敏. 基于轻量级改进网络的无人机航拍图像语义分割研究[J]. 微电子学与计算机, 2024, 41(7): 29-36.
LI Wei, YANG Min. Research on semantic segmentation of aerial images based on improved lightweight networks[J]. Microelectronics & computer, 2024, 41(7): 29-36.
[11] CHEN Guangchen, SHI Benjie, ZHANG Yinhui, et al. CGSNet: Cross-consistency guiding semi-supervised semantic segmentation network for remote sensing of plateau lake[J]. Journal of network and computer applications, 2024, 230: 103974.
[12] 董微微. 考虑横向风浪干扰的舰船航向稳定性自动控制算法[J]. 舰船科学技术, 2024, 46(9): 143-146.
DONG Weiwei. Automatic control algorithm for ship heading stability considering lateral wind and wave interference[J]. Ship science and technology, 2024, 46(9): 143-146.
[13] 康玥, 杨军. 可学习动态分组卷积神经网络的大规模点云分割[J]. 计算机工程与应用, 2024, 60(10): 217-226.
KANG Yue, YANG Jun. Large-scale point cloud segmentation by learnable dynamic grouping convolutional neural network[J]. Computer engineering and applications, 2024, 60(10): 217-226.
[14] 苗壮, 崔浩然, 张启阳, 等. 基于对比学习的大气湍流退化图像复原方法[J]. 计算机科学, 2025, 52(5): 171-178.
MIAO Zhuang, CUI Haoran, ZHANG Qiyang, et al. Restoration of atmospheric turbulence-degraded images based on contrastive learning[J]. Computer science, 2025, 52(5): 171-178.
[15] 王潇棠, 闫河, 刘建骐, 等. 一种边缘梯度插值的双分支deeplabv3+语义分割模型[J]. 智能系统学报, 2023, 18(3): 604-612.
WANG Xiaotang, YAN He, LIU Jianqi, et al. A new deeplabv3+ semantic segmentation model of edge gradient interpolation with double branch structure[J]. CAAI transactions on intelligent systems, 2023, 18(3): 604-612.
[16] 华怡坦, 黄影平, 过文昊. 基于CNN和Transformer点云图像融合的道路检测[J]. 应用科学学报, 2024, 42(4): 695-708.
HUA Yitan, HUANG Yingping, GUO Wenhao. Fusion of point-cloud and image for road segmentation using CNN and Transformer[J]. Journal of applied sciences, 2024, 42(4): 695-708.
[17] HAN Shuo, GAO Lisha, WU Yue, et al. SiamSMN: Siamese cross-modality fusion network for object tracking[J]. Information, 2024, 15(7): 418.
[18] CHEN Chen, LI Ying, WANG Tianqi. Real-time tracking and berthing aid system with occlusion handling based on LiDAR[J]. Ocean engineering, 2023, 288: 115929.
[19] ?UST L, KRISTAN M. Learning with weak annotations for robust maritime obstacle detection[J]. Sensors, 2022, 22(23): 9139.
[20] CHEN L C, ZHU Yukun, PAPANDREOU G, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[C]// Computer Vision–ECCV 2018. Cham: Springer International Publishing, 2018: 833-851.
[21] J?RGES C, BERKENBRINK C, GOTTSCHALK H, et al. Spatial ocean wave height prediction with CNN mixed-data deep neural networks using random field simulated bathymetry[J]. Ocean engineering, 2023, 271: 113699.
[22] 张远, 赵建国. 基于边缘特征点筛选的被遮挡盲道检测算法[J]. 科学技术与工程, 2021, 21(34): 14657-14664.
ZHANG Yuan, ZHAO Jianguo. Algorithm for occluded blind track detection based on edge feature points screening[J]. Science technology and engineering, 2021, 21(34): 14657-14664.
[23] SHI Qingwu, YIN Shoulin, WANG Kun, et al. Multichannel convolutional neural network-based fuzzy active contour model for medical image segmentation[J]. Evolving systems, 2022, 13(4): 535-549.
[24] 邵凯, 王明政, 王光宇. 基于Transformer的多尺度遥感语义分割网络[J]. 智能系统学报, 2024, 19(4): 920-929.
SHAO Kai, WANG Mingzheng, WANG Guangyu. Transformer-based multiscale remote sensing semantic segmentation network[J]. CAAI transactions on intelligent systems, 2024, 19(4): 920-929.
[25] 王卓, 瞿绍军. 基于注意力机制与多尺度池化的实时语义分割网络[J]. 计算机工程, 2023, 49(10): 222-229,238.
WANG Zhuo, QU Shaojun. Real-time semantic segmentation network based on attention mechanism and multi-scale pooling[J]. Computer engineering, 2023, 49(10): 222-229,238.
[26] 赵为平, 陈雨, 项松, 等. 基于改进的DeepLabv3+图像语义分割算法研究[J]. 系统仿真学报, 2023, 35(11): 2333-2344.
ZHAO Weiping, CHEN Yu, XIANG Song, et al. Image semantic segmentation algorithm based on improved DeepLabv3+[J]. Journal of system simulation, 2023, 35(11): 2333-2344.
[27] 刘佳榆, 余华, 徐舒. 基于通道特征选择的加权响应目标跟踪[J]. 计算机仿真, 2022, 39(12): 253-258.
LIU Jiayu, YU Hua, XU Shu. Weighted response object tracking based on channel feature selection[J]. Computer simulation, 2022, 39(12): 253-258.
[28] 方豪, 伍鹏, 谢凯, 等. 基于注意力机制的道路环境语义分割算法[J]. 计算机仿真, 2023, 40(3): 122-128.
FANG Hao, WU Peng, XIE Kai, et al. Semantic segmentation algorithm of road environment based on attention mechanism[J]. Computer simulation, 2023, 40(3): 122-128.
[29] 吕悦, 张永忠. 基于图像语义分割的船舶水线检测[J]. 计算机仿真, 2022, 39(5): 216-219, 233.
LYU Yue, ZHANG Yongzhong. Ship waterline detection based on image semantic segmentation[J]. Computer simulation, 2022, 39(5): 216-219, 233.
[30] 周阳, 叶钦, 杨斌, 等. 海湾台风浪特性及其对船舶的影响[J]. 中国造船, 2020, 61(3): 216-227.
ZHOU Yang, YE Qin, YANG Bin, et al. Wave characteristics in the bay during typhoons and its effect on mooring ships[J]. Shipbuilding of China, 2020, 61(3): 216-227.
[31] 柯熙政, 李鑫. 激光束光斑漂移反演大气湍流各向异性实验研究[J]. 光电工程, 2024, 51(4): 42-53.
KE Xizheng, LI Xin. Experimental study on inversion of atmospheric turbulence anisotropy by laser beam spot wander[J]. Opto-electronic engineering, 2024, 51(4): 42-53.
[32] STEPHENS G L, HAKUBA M Z, KATO S, et al. The changing nature of Earth’s reflected sunlight[J]. Proceedings of the Royal Society A: mathematical, physical and engineering sciences, 2022, 478(2263): 20220053.
[33] TER?EK M, ?UST L, KRISTAN M. eWaSR-an embedded-compute-ready maritime obstacle detection network[J]. Sensors (basel), 2023, 23(12): 5386.
[34] LO W W, YANG Xu, WANG Yapeng. An xception convolutional neural network for malware classification with transfer learning[C]//2019 10th IFIP International Conference on New Technologies, Mobility and Security. CANARY ISLANDS: IEEE, 2019: 1-5.
[35] HOWARD A G, ZHU Menglong, CHEN Bo, et al. MobileNets: efficient convolutional neural networks for mobile vision applications[EB/OL]. (2017-04-17)[2024-05-15]. https://arxiv.org/abs/1704.04861v1.
[36] SANDLER M, HOWARD A, ZHU Menglong , et al. Inverted residuals and linear bottlenecks: Mobile networks for classification, detection and segmentation[EB/OL]//(2018-04-02)[2024-05-05]. https://arxiv.org/abs/1801.04381.
[37] SANDLER M, HOWARD A, ZHU Menglong, et al. MobileNetV2: inverted residuals and linear bottlenecks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 4510-4520.
[38] DAI Yimian, GIESEKE F, OEHMCKE S, et al. Attentional feature fusion[C]//2021 IEEE Winter Conference on Applications of Computer Vision. Waikoloa: IEEE, 2021: 3560-3569.
[39] FU Jun, LIU Jing, JIANG Jie, et al. Scene segmentation with dual relation-aware attention network[J]. IEEE transactions on neural networks and learning systems, 2021, 32(6): 2547-2560.
[40] HOU Qibin, ZHANG Li, CHENG Mingming, et al. Strip pooling: rethinking spatial pooling for scene parsing[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 4003-4012.
[41] EVERINGHAM M, ESLAMI S M A, VAN GOOL L, et al. The pascal visual object classes challenge: a retrospective[J]. International journal of computer vision, 2015, 111: 98-136.
[42] LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft coco: Common objects in context[C]//Computer Vision–ECCV 2014: 13th European Conference. Zurich: Springer International Publishing, 2014: 740-755.
- 相似文献/References:
-
[1]吴涛,董肖莉,孟伟,等.基于语义分割的简洁线条肖像画生成方法[J].智能系统学报,2021,16(1):134.[doi:10.11992/tis.202101003]
WU Tao,DONG Xiaoli,MENG Wei,et al.Concise line portrait generation method based on semantic segmentation[J].CAAI Transactions on Intelligent Systems,2021,16():134.[doi:10.11992/tis.202101003]
[2]王兴武,雷涛,王营博,等.基于多模态互补特征学习的遥感影像语义分割[J].智能系统学报,2022,17(6):1123.[doi:10.11992/tis.202201025]
WANG Xingwu,LEI Tao,WANG Yingbo,et al.Semantic segmentation of remote sensing image based on multimodal complementary feature learning[J].CAAI Transactions on Intelligent Systems,2022,17():1123.[doi:10.11992/tis.202201025]
[3]王潇棠,闫河,刘建骐,等.一种边缘梯度插值的双分支deeplabv3+语义分割模型[J].智能系统学报,2023,18(3):604.[doi:10.11992/tis.202111023]
WANG Xiaotang,YAN He,LIU Jianqi,et al.A new deeplabv3+ semantic segmentation model of edge gradient interpolation with double branch structure[J].CAAI Transactions on Intelligent Systems,2023,18():604.[doi:10.11992/tis.202111023]
[4]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.[doi:10.11992/tis.202210039]
SUN Yuxin,SU Li,CHEN Yusheng,et al.SOLOA ship instance segmentation algorithm based on attention[J].CAAI Transactions on Intelligent Systems,2023,18():1197.[doi:10.11992/tis.202210039]
[5]翟永杰,王家豪,张鑫,等.基于语义分割视觉伺服的种苗自动夹取系统设计[J].智能系统学报,2023,18(6):1259.[doi:10.11992/tis.202212026]
ZHAI Yongjie,WANG Jiahao,ZHANG Xin,et al.Design of automatic picking system for seedlings based on semantic segmentation visual servo[J].CAAI Transactions on Intelligent Systems,2023,18():1259.[doi:10.11992/tis.202212026]
[6]庞荣,杨燕,冷雄进,等.基于双分支点流语义先验的路面病害分割模型[J].智能系统学报,2024,19(1):153.[doi:10.11992/tis.202306037]
PANG Rong,YANG Yan,LENG Xiongjin,et al.Segmentation model of pavement diseases based on semantic priori of double-branched point flow[J].CAAI Transactions on Intelligent Systems,2024,19():153.[doi:10.11992/tis.202306037]
[7]邵凯,王明政,王光宇.基于Transformer的多尺度遥感语义分割网络[J].智能系统学报,2024,19(4):920.[doi:10.11992/tis.202304026]
SHAO Kai,WANG Mingzheng,WANG Guangyu.Transformer-based multiscale remote sensing semantic segmentation network[J].CAAI Transactions on Intelligent Systems,2024,19():920.[doi:10.11992/tis.202304026]
[8]赵振兵,郭广学,王艺衡,等.融合边缘感知与统计纹理知识的输电线路金具锈蚀检测[J].智能系统学报,2024,19(5):1228.[doi:10.11992/tis.202306009]
ZHAO Zhenbing,GUO Guangxue,WANG Yiheng,et al.Rust detection in transmission line fittings via fusion of edge perception and statistical texture knowledge[J].CAAI Transactions on Intelligent Systems,2024,19():1228.[doi:10.11992/tis.202306009]
[9]杨宇宇,杨霄,潘在宇,等.基于原型引导与自适应特征融合的域适应语义分割[J].智能系统学报,2025,20(1):150.[doi:10.11992/tis.202403010]
YANG Yuyu,YANG Xiao,PAN Zaiyu,et al.Domain adaptive semantic segmentation based on prototype-guided and adaptive feature fusion[J].CAAI Transactions on Intelligent Systems,2025,20():150.[doi:10.11992/tis.202403010]
备注/Memo
收稿日期:2024-5-15。
基金项目:国家自然科学基金项目(62033010,62303233)；江苏省高校青蓝工程项目(R2023Q07).
作者简介:沈昊，硕士研究生，主要研究方向为计算机视觉。E-mail：hshen_nuist@163.com。;葛泉波，教授，中国自动化学会青年工作委员会副主任委员、中国人工智能学会自主无人系统专业委员会副主任委员、中国自动化学会人工智能与机器人教育专业委员会副秘书长、中国自动化学会混合智能专业委员会副秘书长等。江苏省“青蓝工程”中青年学术带头人、浙江省万人计划“青年拔尖”入选者、中国自动化学会第四届青年科学家奖、浙江省杰出青年科学基金获得者等。主要研究方向为状态估计与信息融合、自主智能无人系统、飞行器测试数据分析和电力IOT技术。主持国家级科研项目12项，省部级科研项目7项，其余类别科研项目10项。发表学术论文92篇，制定专利/标准76项。E-mail：quanboge@163.com。;吴高峰，讲师，主要研究方向为航空火力控制、无人机任务规划与控制、机器学习及其在无人系统中应用。E-mail：wgf@nuist.edu.cn。
通讯作者:葛泉波. E-mail：quanboge@163.com
更新日期/Last Update:
1900-01-01