海洋涡旋智能检测研究进展

doi:10.3969/j.issn.1001-909X.2024.03.003

海洋学研究 ›› 2024, Vol. 42 ›› Issue (3): 38-50.DOI: 10.3969/j.issn.1001-909X.2024.03.003

海洋涡旋智能检测研究进展

徐广珺¹^,²(), 施宇诚¹, 余洋³, 谢华荣⁴, 谢文鸿⁵, 刘婧媛¹, 林夏艳⁶, 刘宇⁶, 董昌明²^,⁴^,^*()

1.广东海洋大学电子与信息工程学院,广东湛江 524088
2.南方海洋科学与工程广东省实验室(珠海),广东珠海 519000
3.福建省气象台, 福建福州 350007
4.南京信息工程大学海洋科学学院,江苏南京 210044
5.南京星遥科技有限公司,江苏南京 210012
6.浙江海洋大学海洋科学与技术学院,浙江舟山 316022

收稿日期:2023-12-30 修回日期:2024-08-12 出版日期:2024-09-15 发布日期:2024-11-25
通讯作者: ^*董昌明(1967—),男,教授,主要从事人工智能海洋学研究,E-mail:cmdong@nuist.edu.cn。
作者简介:徐广珺(1987—),男,江苏省南京市人,主要从事人工智能海洋学、卫星海洋动力学研究,E-mail:gjxu_gdou@yeah.net。
基金资助:
国家重点研发计划项目(2023YFC3008200);南方海洋科学与工程广东省实验室(珠海)自主科研项目(SML2020SP007);热带海洋环境国家重点实验室(中国科学院南海海洋研究所)开放课题(LTO2319)

Recent developments in AI-based oceanic eddy identification

XU Guangjun¹^,²(), SHI Yucheng¹, YU Yang³, XIE Huarong⁴, XIE Wenhong⁵, LIU Jingyuan¹, LIN Xiayan⁶, LIU Yu⁶, DONG Changming²^,⁴^,^*()

1. School of Electronics and Information Engineering, Guangdong Ocean University, Zhanjiang 524088, China
2. Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai 519000, China
3. Fujian Provincial Meteorological Observatory, Fuzhou 350007, China
4. School of Marine Sciences, Nanjing University of Information Science and Technology, Nanjing 210044, China
5. Nanjing Xingyao Technology Co., LTD, Nanjing 210012, China
6. Marine Science and Technology College, Zhejiang Ocean University, Zhoushan 316022, China

Received:2023-12-30 Revised:2024-08-12 Online:2024-09-15 Published:2024-11-25

摘要/Abstract

摘要：

海洋涡旋是一种常见的海洋现象,在全球海洋物质和能量的输运中起着重要作用。随着海洋研究技术手段的不断提升,各类海洋涡旋检测方法应运而生。传统涡旋检测方法应用广泛,但其过度依赖于专家经验设置阈值和持续的人工干预,存在检测误差较大、工作效率低以及全球普适性差等问题,难以适应复杂多变的海洋环境。当前人工智能快速发展,其在海洋涡旋智能检测中能够自动、快速地提取图像深层特征,有效解决海洋现象特征相似度高、几何差异大的问题。该文立足于当前海洋涡旋智能检测的发展现状,从编码器-解码器结构、全卷积神经网络、多尺度上下文方法和注意力机制等方面回顾了不同深度学习方法在海洋涡旋智能检测中的应用,以期为海洋涡旋研究提供一些启示和参考。

关键词: 海洋涡旋, 人工智能, 特征检测, 深度学习, 编码器-解码器结构, 全卷积神经网络, 多尺度上下文方法, 注意力机制

Abstract:

Ocean eddies are prevalent oceanic phenomenon that play a crucial role in the global transportation of oceanic materials and energy. Although traditional methods for detecting ocean eddies are widely used, they suffer from significant drawbacks such as excessive reliance on expert-set thresholds, continuous manual intervention, large detection errors, low efficiency, and poor global applicability, making it difficult to adapt to the complex and variable marine environment. Currently, the rapid development of artificial intelligence (AI) presents a promising solution for the intelligent detection of ocean eddies. AI can automatically and rapidly extract deep features from images, effectively address the challenges posed by the high similarity in oceanic phenomenon features and significant geometric variability. This paper provides an overview of AI-based oceanic eddy identification methods based on different deep learning methods, focuses on coder-decoder structure, fully convolutional neural network, multi-scale context method and attention mechanism, and aims to provide valuable insights and references for future ocean eddy research.

Key words: oceanic eddy, artificial intelligence, feature detection, deep learning, coder-decoder structure, fully convolutional neural network, multi-scale context method, attention mechanism

中图分类号:

P731.2

徐广珺, 施宇诚, 余洋, 谢华荣, 谢文鸿, 刘婧媛, 林夏艳, 刘宇, 董昌明. 海洋涡旋智能检测研究进展[J]. 海洋学研究, 2024, 42(3): 38-50.

XU Guangjun, SHI Yucheng, YU Yang, XIE Huarong, XIE Wenhong, LIU Jingyuan, LIN Xiayan, LIU Yu, DONG Changming. Recent developments in AI-based oceanic eddy identification[J]. Journal of Marine Sciences, 2024, 42(3): 38-50.

图/表 7

图1 中尺度涡旋在三种不同图像中的表征

Fig.1 Representation of mesoscale eddies in three different images

图2 基于CNN的海洋涡旋检测模型[26] (图中蓝色表示卷积层、批归一化层和ReLU激活层;绿色表示池化层;橙色表示上采样层;红色表示softmax激活层。)

Fig.2 Oceanic eddy detection model based on CNN[26] (Blue: convolution, batch normalization and ReLU-activation layers; green: pooling layers; orange: up-sampling layers; red: softmax-activation layer.)

图3 基于深度学习的异常涡旋检测DL-AED模型的示意图及异常涡旋检测结果 (图件改绘自文献[35]。)

Fig.3 Schematic diagram of the deep learning-based abnormal eddy detection DL-AED model and its output (Figure is redrawn from reference [35].)

图4 基于高频地波雷达智能检测的海洋涡旋[41]

Fig.4 Intelligent detection of oceanic eddies based on HFR[41]

图5 金字塔场景解析网络模型示意图 (图件改绘自文献[42]。)

Fig.5 Schematic diagram of the pyramid scene parsing network (PSPNet) (Figure is redrawn from reference [42].)

图6 上下文与边缘关联网络的示意图 (图件改绘自文献[48]。)

Fig.6 Schematic diagram of context and edge association network (CEA-Net) (Figure is redrawn from reference [48].)

图7 双重注意力网络的示意图 (图件改绘自文献[49]。)

Fig.7 Schematic diagram of the dual attention network (Danet) (Figure is redrawn from reference [49].)

参考文献 52

[1]	MORROW R, LE TRAON P Y. Recent advances in observing mesoscale ocean dynamics with satellite altimetry[J]. Advances in Space Research, 2012, 50(8): 1062-1076.
[2]	WANG H Z, LIU D, ZHANG W M, et al. Characterizing the capability of mesoscale eddies to carry drifters in the northwest Pacific[J]. Journal of Oceanology and Limnology, 2020, 38(6): 1711-1728.
[3]	DONG C M, MCWILLIAMS J C, LIU Y, et al. Global heat and salt transports by eddy movement[J]. Nature Commu-nications, 2014, 5: 3294.
[4]	JI J L, DONG C M, ZHANG B, et al. An oceanic eddy statistical comparison using multiple observational data in the Kuroshio extension region[J]. Acta Oceanologica Sinica, 2017, 36(3): 1-7.
[5]	FRENGER I, MÜNNICH M, GRUBER N, et al. Southern Ocean eddy phenomenology[J]. Journal of Geophysical Research: Oceans, 2015, 120(11): 7413-7449.
[6]	CHELTON D B, SCHLAX M G, SAMELSON R M, et al. Global observations of large oceanic eddies[J]. Geophysical Research Letters, 2007, 34(15): L15606.
[7]	董昌明. 海洋涡旋探测与分析[M]. 北京: 科学出版社, 2015.
	DONG C M. Oceanic eddy detection and analysis[M]. Beijing: Science Press, 2015.
[8]	FRENGER I, GRUBER N, KNUTTI R, et al. Imprint of Southern Ocean eddies on winds, clouds and rainfall[J]. Nature Geoscience, 2013, 6: 608-612.
[9]	GAUBE P, CHELTON D B, SAMELSON R M, et al. Satellite observations of mesoscale eddy-induced Ekman pumping[J]. Journal of Physical Oceanography, 2015, 45(1): 104-132.
[10]	GAUBE P, BARCELÓ C, MCGILLICUDDY D J Jr, et al. The use of mesoscale eddies by juvenile loggerhead sea turtles (Caretta caretta) in the southwestern Atlantic[J]. PLoS One, 2017, 12(3): e0172839.
[11]	李佳讯, 张韧, 陈奕德, 等. 海洋中尺度涡建模及其在水声传播影响研究中的应用[J]. 海洋通报, 2011, 30(1):37-46.
	LI J X, ZHANG R, CHEN Y D, et al. Ocean mesoscale eddy modeling and its application in studying the effect on underwater acoustic propagation[J]. Marine Science Bulletin, 2011, 30(1): 37-46.
[12]	DONG C M, LIU Y, LUMPKIN R, et al. A scheme to identify loops from trajectories of oceanic surface drifters: An application in the Kuroshio extension region[J]. Journal of Atmospheric and Oceanic Technology, 2011, 28(9): 1167-1176.
[13]	NENCIOLI F, DONG C M, DICKEY T, et al. A vector geometry-based eddy detection algorithm and its application to a high-resolution numerical model product and high-frequency radar surface velocities in the southern California bight[J]. Journal of Atmospheric and Oceanic Technology, 2010, 27(3): 564-579.
[14]	ZHANG C H, LI H L, LIU S T, et al. Automatic detection of oceanic eddies in reanalyzed SST images and its application in the East China Sea[J]. Science China Earth Sciences, 2015, 58(12): 2249-2259.
[15]	SUN W J, LIU Y, CHEN G X, et al. Three-dimensional properties of mesoscale cyclonic warm-core and anticyclonic cold-core eddies in the South China Sea[J]. Acta Oceanologica Sinica, 2021, 40(10): 17-29.
[16]	XU G J, YANG J S, DONG C M, et al. Statistical study of submesoscale eddies identified from synthetic aperture radar images in the Luzon Strait and adjacent seas[J]. International Journal of Remote Sensing, 2015, 36(18): 4621-4631.
[17]	JI Y X, XU G J, DONG C M, et al. Submesoscale eddies in the East China Sea detected from SAR images[J]. Acta Oceanologica Sinica, 2021, 40(3): 18-26.
[18]	OKUBO A. Horizontal dispersion of floatable particles in the vicinity of velocity singularities such as convergences[J]. Deep Sea Research and Oceanographic Abstracts, 1970, 17(3): 445-454.
[19]	WEISS J. The dynamics of enstrophy transfer in two-dimensional hydrodynamics[J]. Physica D: Nonlinear Phenomena, 1991, 48(2/3): 273-294.
[20]	HUNT J C R. Vorticity and vortex dynamics in complex turbulent flows[J]. Transactions of the Canadian Society for Mechanical Engineering, 1987, 11(1): 21-35.
[21]	LIU C Q, WANG Y Q, YANG Y, et al. New omega vortex identification method[J]. Science China Physics, Mechanics & Astronomy, 2016, 59(8): 684711.
[22]	CHONG M S, PERRY A E, CANTWELL B J. A general classification of three-dimensional flow fields[J]. Physics of Fluids A: Fluid Dynamics, 1990, 2(5): 765-777.
[23]	JEONG J, HUSSAIN F. On the identification of a vortex[J]. Journal of Fluid Mechanics, 1995, 285: 69-94.
[24]	DOGLIOLI A M, BLANKE B, SPEICH S, et al. Tracking coherent structures in a regional ocean model with wavelet analysis: Application to Cape Basin eddies[J]. Journal of Geophysical Research: Oceans, 2007, 112(C5): C05043.
[25]	CHAIGNEAU A, GIZOLME A, GRADOS C. Mesoscale eddies off Peru in altimeter records: Identification algorithms and eddy spatio-temporal patterns[J]. Progress in Oceano-graphy, 2008, 79(2/3/4): 106-119.
[26]	FRANZ K, ROSCHER R, MILIOTO A, et al. Ocean eddy identification and tracking using neural networks[C]//2018 IEEE International Geoscience and Remote Sensing Sympo-sium(IGARSS). IEEE, 2018: 6887-6890.
[27]	LGUENSAT R, SUN M, FABLET R, et al. EddyNet: A deep neural network for pixel-wise classification of oceanic eddies[C]//2018 IEEE International Geoscience and Remote Sensing Symposium(IGARSS). IEEE, 2018: 1764-1767.
[28]	LI X F, LIU B, ZHENG G, et al. Deep-learning-based information mining from ocean remote-sensing imagery[J]. National Science Review, 2020, 7(10): 1584-1605. DOI PMID
[29]	DUO Z J, WANG W K, WANG H Z. Oceanic mesoscale eddy detection method based on deep learning[J]. Remote Sensing, 2019, 11(16): 1921.
[30]	SANTANA O J, HERNÁNDEZ-SOSA D, SMITH R N. Oceanic mesoscale eddy detection and convolutional neural network complexity[J]. International Journal of Applied Earth Observation and Geoinformation, 2022, 113: 102973.
[31]	张盟, 杨玉婷, 孙鑫, 等. 基于深度卷积网络的海洋涡旋检测模型[J]. 南京航空航天大学学报, 2020, 52(5):708-713.
	ZHANG M, YANG Y T, SUN X, et al. Ocean eddy detection model based on deep convolution neural network[J]. Journal of Nanjing University of Aeronautics & Astronautics, 2020, 52(5): 708-713.
[32]	刘启明, 杨树国, 赵莉. 基于深度卷积神经网络的海洋多目标涡旋检测方法[J]. 青岛科技大学学报:自然科学版, 2022, 43(4):120-126.
	LIU Q M, YANG S G, ZHAO L. Ocean multi-eddy detection method based on deep convolution neural network[J]. Journal of Qingdao University of Science and Technology: Natural Science Edition, 2022, 43(4): 120-126.
[33]	LIU Y J, LI X F, REN Y B. A deep learning model for oceanic mesoscale eddy detection based on multi-source remote sensing imagery[C]//2020 IEEE International Geoscience and Remote Sensing Symposium(IGARSS). IEEE, 2020: 6762-6765.
[34]	沈飙, 陈扬, 杨琛, 等. 海洋科学中尺度涡的计算机视觉检测和分析方法[J]. 数据与计算发展前沿, 2020, 2(6):30-41.
	SHEN B, CHEN Y, YANG C, et al. Computer vision detection and analysis of mesoscale eddies in marine science[J]. Frontiers of Data & Computing, 2020, 2(6): 30-41.
[35]	LIU Y J, ZHENG Q A, LI X F. Characteristics of global ocean abnormal mesoscale eddies derived from the fusion of sea surface height and temperature data by deep learning[J]. Geophysical Research Letters, 2021, 48(17): e94772.
[36]	CAO L J, ZHANG D J, ZHANG X F, et al. Detection and identification of mesoscale eddies in the South China Sea based on an artificial neural network model—YOLOF and remotely sensed data[J]. Remote Sensing, 2022, 14(21): 5411.
[37]	YAN Z F, CHONG J S, ZHAO Y W, et al. Multifeature fusion neural network for oceanic phenomena detection in SAR images[J]. Sensors, 2019, 20(1): 210.
[38]	DU Y L, SONG W, HE Q, et al. Deep learning with multi-scale feature fusion in remote sensing for automatic oceanic eddy detection[J]. Information Fusion, 2019, 49: 89-99.
[39]	ZHANG D, GADE M, ZHANG J W. SAR eddy detection using mask-RCNN and edge enhancement[C]//2020 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), September 26-October 2, 2020, Waikoloa, HI, USA. IEEE, 2020: 1604-1607.
[40]	贾翊文, 荆文龙, 杨骥, 等. 基于深度学习的SAR影像海洋涡旋检测算法对比分析[J]. 海洋科学进展, 2024, 42(1):137-148.
	JIA Y W, JING W L, YANG J, et al. Comparative analysis of ocean eddy detection algorithms based on deep learning in SAR images[J]. Advances in Marine Science, 2024, 42(1): 137-148.
[41]	LIU F Y, ZHOU H, HUANG W M, et al. Cross-domain submesoscale eddy detection neural network for HF radar[J]. Remote Sensing, 2021, 13(13): 2441.
[42]	XU G J, CHENG C, YANG W X, et al. Oceanic eddy identification using an AI scheme[J]. Remote Sensing, 2019, 11: 1349.
[43]	XU G J, XIE W H, DONG C M, et al. Application of three deep learning schemes into oceanic eddy detection[J]. Frontiers in Marine Science, 2021, 8: 672334.
[44]	HANG R L, LI G, XUE M, et al. Identifying oceanic eddy with an edge-enhanced multiscale convolutional network[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 15: 9198-9207.
[45]	ZHAO Y X, FAN Z L, LI H T, et al. SymmetricNet: End-to-end mesoscale eddy detection with multi-modal data fusion[J]. Frontiers in Marine Science, 2023, 10: 1174818.
[46]	杜艳玲, 王丽丽, 黄冬梅, 等. 融合密集特征金字塔的改进R²CNN海洋涡旋自动检测[J]. 智能系统学报, 2023, 18(2):341-351.
	DU Y L, WANG L L, HUANG D M, et al. Improved R²CNN ocean eddy automatic detection with a dense feature pyramid[J]. CAAI Transactions on Intelligent Systems, 2023, 18(2): 341-351.
[47]	FAN Z L, ZHONG G Q. SymmetricNet: A mesoscale eddy detection method based on multivariate fusion data[Z/OL]. arXiv, 2019: 1909. 13411. https://arxiv.org/abs/1909.13411v1.
[48]	XIA L H, CHEN G, CHEN X Y, et al. Submesoscale oceanic eddy detection in SAR images using context and edge association network[J]. Frontiers in Marine Science, 2022, 9: 1023624.
[49]	ZHANG Y Y, LIU N, ZHANG Z Y, et al. Detection of Bering Sea slope mesoscale eddies derived from satellite altimetry data by an attention network[J]. Remote Sensing, 2022, 14(19): 4974.
[50]	SAIDA S J, SAHOO S P, ARI S. Deep convolution neural network based semantic segmentation for ocean eddy detection[J]. Expert Systems with Applications, 2023, 219: 119646.
[51]	ZHAO N, HUANG B X, YANG J, et al. Oceanic eddy identification using pyramid split attention U-net with remote sensing imagery[J]. IEEE Geoscience and Remote Sensing Letters, 2023, 20: 1-5.
[52]	杜艳玲, 吴天宇, 陈括, 等. 融合上下文和注意力的海洋涡旋小目标检测[J]. 中国图象图形学报, 2023, 28(11):3509-3519.
	DU Y L, WU T Y, CHEN K, et al. Small object detection for ocean eddies using contextual information and attention mechanism[J]. Journal of Image and Graphics, 2023, 28(11): 3509-3519.

海洋涡旋智能检测研究进展

Recent developments in AI-based oceanic eddy identification

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 52

相关文章 5

编辑推荐

Metrics

[1]	邵志远, 赵杰臣, 解龙翔, 牟芳如, 肖静, 刘敏君, 陈雪婧. 基于深度学习和Sentinel-1卫星影像的北极海冰分类精度和影响因素[J]. 海洋学研究, 2024, 42(3): 119-130.
[2]	董昌明, 王子韵, 谢华荣, 徐广珺, 韩国庆, 周书逸, 谢文鸿, 沈向宇, 韩磊. 人工智能海洋学发展前景[J]. 海洋学研究, 2024, 42(3): 2-27.
[3]	陆钰婷, 郭文康, 丁骏, 王林峰, 李晓辉, 王久珂. 人工智能海浪预报的发展与挑战[J]. 海洋学研究, 2024, 42(3): 28-37.
[4]	邸雅楠, 赵若轩, 徐建洲. 基于深度学习的生物组织病理图像分析在海洋监测中的发展潜力及案例分析[J]. 海洋学研究, 2024, 42(3): 64-74.
[5]	金阳, 韩磊, 金梅兵, 董昌明. 基于ConvLSTM的中国东南沿海波浪智能预报和评估[J]. 海洋学研究, 2024, 42(3): 88-98.