Article(id=1212062584971334532, tenantId=1146029695717560320, journalId=1149651085930835976, issueId=1212062580651201329, articleNumber=null, orderNo=null, doi=10.12284/hyxb2023151, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1676390400000, receivedDateStr=2023-02-15, revisedDate=1686585600000, revisedDateStr=2023-06-13, acceptedDate=null, acceptedDateStr=null, onlineDate=1766907839291, onlineDateStr=2025-12-28, pubDate=1696089600000, pubDateStr=2023-10-01, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1766907839291, onlineIssueDateStr=2025-12-28, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1766907839291, creator=13701087609, updateTime=1766907839291, updator=13701087609, issue=Issue{id=1212062580651201329, tenantId=1146029695717560320, journalId=1149651085930835976, year='2023', volume='45', issue='10', pageStart='1', pageEnd='194', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1766907838261, creator=13701087609, updateTime=1766924731029, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1212133434105918266, tenantId=1146029695717560320, journalId=1149651085930835976, issueId=1212062580651201329, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1212133434105918267, tenantId=1146029695717560320, journalId=1149651085930835976, issueId=1212062580651201329, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=168, endPage=182, ext={EN=ArticleExt(id=1212062585294295959, articleId=1212062584971334532, tenantId=1146029695717560320, journalId=1149651085930835976, language=EN, title=Automatic extraction of green tide in areas with clouds or solar flares in HY-1C/D CZI multispectral images, columnId=1194652705852465724, journalTitle=Haiyang Xuebao, columnName=Article, runingTitle=null, highlight=null, articleAbstract=

Multispectral images are greatly affected by factors such as clouds, fog, and solar flares, which makes it difficult to automatically extract high-precision green tides under complex weather conditions. Based on the multi-spectral images of my country’s HY-1C/D satellite CZI payload, using data mining technology to explore the difference in data distribution between green tide areas and non-green tide areas, we propose a high-precision and fully automatic green tide extraction method , which can be applied to HY-1C/D CZI sensor data. First of all, the thick cloud area is removed by preliminary extraction rules to achieve preliminary classification. Then, the correctly classified green tide samples and non-green tide samples were used as positive and negative samples respectively, and these samples were used as experimental data to train the decision tree model, and the automatic extraction rules of green tide were obtained according to the model. Finally, 5 strategies for correcting misclassifications were designed to achieve fully automatic extraction of green tides. In order to verify the effectiveness of the method, we collected 25 images of the green tide outbreak period in the Yellow Sea in 2021 for automatic detection experiments, and compared the experimental results with traditional index methods (NDVI, VB-FAH) and deep learning methods (ResNet50, U-Net). The results showed that the method outperformed other methods in terms of accuracy, Kappa coefficient, F1-Score, and MIoU. The accuracy of green tide extraction was higher in areas with thick clouds, thin clouds, cloudless clouds, cloud spots, and flares.

, correspAuthors=Changying Wang, authorNote=null, correspAuthorsNote=null, copyrightStatement=Copyright © 2023 Pratacultural Science. All rights reserved., copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=Ke Wu, Changying Wang, Rui Huang, Huawei Li), CN=ArticleExt(id=1212062588532297835, articleId=1212062584971334532, tenantId=1146029695717560320, journalId=1149651085930835976, language=CN, title=HY-1C/D CZI多光谱影像云覆盖与耀斑区域绿潮自动提取方法研究, columnId=1149698756456657529, journalTitle=海洋学报, columnName=论文, runingTitle=null, highlight=null, articleAbstract=

针对多光谱影像受云、雾、太阳耀斑等因素的影响,难以实现高精度的绿潮自动提取的问题,本文以我国的HY-1C/D卫星CZI载荷多光谱影像为数据源,采用数据挖掘技术,通过探索绿潮区域与非绿潮区域的光谱分布差异,提出一种适用于HY-1C/D CZI影像的高精度、全自动绿潮提取方法。首先,分析有云区域和无云区域样本的光谱差异,给出厚云去除规则;其次,选取绿潮和非绿潮区域的样本,采用决策树算法生成绿潮提取规则;然后,针对薄云和厚云边界区域常常会出现误检绿潮的问题,设计了5种错误类别修正策略。为验证方法的有效性,收集2021年黄海区域绿潮暴发周期内的25景HY-1C/D CZI影像,开展绿潮自动检测实验。结果表明,与传统的NDVI方法、VB-FAH方法等指数方法以及ResNet50、U-Net等深度学习方法相比,本文方法在准确度、Kappa系数、F1-Score和MIoU等指标上均优于其他方法,而且能够实现在厚云、薄云、无云、云斑和耀斑区域复杂情况下的绿潮的高精度自动提取。

, correspAuthors=王常颖, authorNote=null, correspAuthorsNote=
*王常颖(1980—),副教授,主要从事海洋复杂性与数据挖掘研究。E-mail:
, copyrightStatement=版权所有©《海洋学报》编辑部 2023, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=yYLahUZbeSfLuTV8cg3r9w==, magXml=w9sfGD7pjlqgN2JPu/YHyg==, pdfUrl=null, pdf=TYomHsveDRdKRuRtxnFRXw==, pdfFileSize=7198870, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=GEW8fKObiDvTysWFAgeDtQ==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=2xlov7NRQEyBqEQorNfRlg==, mapNumber=null, authorCompany=null, fund=null, authors=

吴克(1999—),男,河南省濮阳市人,研究方向为遥感大数据。E-mail:

, authorsList=吴克, 王常颖, 黄睿, 李华伟)}, authors=[Author(id=1215325293972210203, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=1962612978@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1215325294060290592, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, authorId=1215325293972210203, language=EN, stringName=Ke Wu, firstName=Ke, middleName=null, lastName=Wu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=null, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1215325294156759592, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, authorId=1215325293972210203, language=CN, stringName=吴克, firstName=克, middleName=null, lastName=吴, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 青岛大学 计算机科学技术学院,山东 青岛, 266071, bio={"content":"

吴克(1999—),男,河南省濮阳市人,研究方向为遥感大数据。E-mail:

"}, bioImg=null, bioContent=

吴克(1999—),男,河南省濮阳市人,研究方向为遥感大数据。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1215325293666025990, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, xref=1, ext=[AuthorCompanyExt(id=1215325293674414599, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, companyId=1215325293666025990, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 青岛大学 计算机科学技术学院,山东 青岛, 266071)])]), Author(id=1215325294240645678, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wcing@qdu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1215325294328726070, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, authorId=1215325294240645678, language=EN, stringName=Changying Wang, firstName=Changying, middleName=null, lastName=Wang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=null, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1215325294421000766, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, authorId=1215325294240645678, language=CN, stringName=王常颖, firstName=常颖, middleName=null, lastName=王, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 青岛大学 计算机科学技术学院,山东 青岛, 266071, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1215325293666025990, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, xref=1, ext=[AuthorCompanyExt(id=1215325293674414599, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, companyId=1215325293666025990, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 青岛大学 计算机科学技术学院,山东 青岛, 266071)])]), Author(id=1215325294504886851, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1215325294609744459, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, authorId=1215325294504886851, language=EN, stringName=Rui Huang, firstName=Rui, middleName=null, lastName=Huang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=null, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1215325294714602068, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, authorId=1215325294504886851, language=CN, stringName=黄睿, firstName=睿, middleName=null, lastName=黄, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 青岛大学 计算机科学技术学院,山东 青岛, 266071, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1215325293666025990, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, xref=1, ext=[AuthorCompanyExt(id=1215325293674414599, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, companyId=1215325293666025990, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 青岛大学 计算机科学技术学院,山东 青岛, 266071)])]), Author(id=1215325294844625501, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1215325294978843238, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, authorId=1215325294844625501, language=EN, stringName=Huawei Li, firstName=Huawei, middleName=null, lastName=Li, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=null, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1215325295071117931, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, authorId=1215325294844625501, language=CN, stringName=李华伟, firstName=华伟, middleName=null, lastName=李, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 青岛大学 计算机科学技术学院,山东 青岛, 266071, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1215325293666025990, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, xref=1, ext=[AuthorCompanyExt(id=1215325293674414599, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, companyId=1215325293666025990, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 青岛大学 计算机科学技术学院,山东 青岛, 266071)])])], keywords=[Keyword(id=1215325295301804660, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, orderNo=1, keyword=HY-1C/D satellite), Keyword(id=1215325295406662265, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, orderNo=2, keyword=green tide extraction), Keyword(id=1215325295494742655, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, orderNo=3, keyword=decision tree), Keyword(id=1215325295587017350, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, orderNo=4, keyword=solar flare), Keyword(id=1215325295696069259, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, orderNo=5, keyword=cloud cover), Keyword(id=1215325295796732561, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, orderNo=1, keyword=HY-1C/D卫星), Keyword(id=1215325297059218069, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, orderNo=2, keyword=绿潮提取), Keyword(id=1215325297155687066, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, orderNo=3, keyword=决策树), Keyword(id=1215325297214407328, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, orderNo=4, keyword=耀斑), Keyword(id=1215325297285710499, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, orderNo=5, keyword=云覆盖)], refs=[Reference(id=1215325303052878722, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=null, journalName=null, refType=null, unstructuredReference=Hiraoka M, Ohno M, Kawaguchi S, et al. Crossing test among floating Ulva thalli forming ‘green tide’ in Japan[J]. Hydrobiologia, 2004, 512(1): 239−245., articleTitle=null, refAbstract=null), Reference(id=1215325303119987590, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=衣立, 张苏平, 殷玉齐. 2009年黄海绿潮浒苔爆发与漂移的水文气象环境[J]. 中国海洋大学学报, 2010, 40(10): 15−23., articleTitle=null, refAbstract=null), Reference(id=1215325303187096457, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=Yi Li, Zhang Suping, Yin Yuqi. Influnce of environmental hydro-meteorological conditions to Enteromorpha prolifera blooms in Yellow Sea, 2009[J]. Periodical of Ocean University of China, 2010, 40(10): 15−23., articleTitle=null, refAbstract=null), Reference(id=1215325303279371147, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=3, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Dongyan, Keesing J K, He Peimin, et al. The world’s largest macroalgal bloom in the Yellow Sea, China: formation and implications[J]. Estuarine, Coastal and Shelf Science, 2013, 129: 2−10., articleTitle=null, refAbstract=null), Reference(id=1215325303354868621, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=Hu Lianbo, Zeng Kan, Hu Chuanmin, et al. On the remote estimation of Ulva prolifera areal coverage and biomass[J]. Remote Sensing of Environment, 2019, 223: 194−207., articleTitle=null, refAbstract=null), Reference(id=1215325303421977489, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=Qi Lin, Hu Chuanmin, Xing Qianguo, et al. Long-term trend of Ulva prolifera blooms in the western Yellow Sea[J]. Harmful Algae, 2016, 58: 35−44., articleTitle=null, refAbstract=null), Reference(id=1215325303501669269, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=Qi Lin, Hu Chuanmin. To what extent can Ulva and Sargassum be detected and separated in satellite imagery?[J]. Harmful Algae, 2021, 103: 102001., articleTitle=null, refAbstract=null), Reference(id=1215325303623304088, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=Hu Lianbo, Hu Chuanmin, He Mingxia. Remote estimation of biomass of Ulva prolifera macroalgae in the Yellow Sea[J]. Remote Sensing of Environment, 2017, 192: 217−227., articleTitle=null, refAbstract=null), Reference(id=1215325303702995868, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=Hu Chuanmin, Li Daqiu, Chen Changsheng, et al. On the recurrent Ulva prolifera blooms in the Yellow Sea and East China Sea[J]. Journal of Geophysical Research: Oceans, 2010, 115(C5): C05017., articleTitle=null, refAbstract=null), Reference(id=1215325303795270556, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=Xing Qianguo, Hu Chuanmin, Tang Danling, et al. World’s largest macroalgal blooms altered phytoplankton biomass in summer in the Yellow Sea: satellite observations[J]. Remote Sensing, 2015, 7(9): 12297−12313., articleTitle=null, refAbstract=null), Reference(id=1215325303929488288, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=Nelson T A, Haberlin K, Nelson A V, et al. Ecological and physiological controls of species composition in green macroalgal blooms[J]. Ecology, 2008, 89(5): 1287−1298., articleTitle=null, refAbstract=null), Reference(id=1215325304034345889, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=Hu Chuanmin, He Mingxia. Origin and offshore extent of floating algae in Olympic sailing area[J]. Eos, Transactions American Geophysical Union, 2008, 89(33): 302−303., articleTitle=null, refAbstract=null), Reference(id=1215325304130814884, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=施英妮, 石立坚, 夏明, 等. HJ-1A/1B星CCD传感器数据在黄东海浒苔监测中的应用[J]. 遥感信息, 2012, 27(2): 47−50., articleTitle=null, refAbstract=null), Reference(id=1215325304248255402, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=Shi Yingni, Shi Lijian, Xia Ming, et al. The application of HJ-1A/1B’s CCD data to entermorpha prolifera monitoring over the Yellow Sea and East Sea[J]. Remote Sensing Information, 2012, 27(2): 47−50., articleTitle=null, refAbstract=null), Reference(id=1215325304348918704, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=Hu Chuanmin. A novel ocean color index to detect floating algae in the global oceans[J]. Remote Sensing of Environment, 2009, 113(10): 2118−2129., articleTitle=null, refAbstract=null), Reference(id=1215325304428610485, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=Son Y B, Min J E, Ryu J H. Detecting massive green algae (Ulva prolifera) blooms in the Yellow Sea and East China Sea using geostationary ocean color imager (GOCI) data[J]. Ocean Science Journal, 2012, 47(3): 359−375., articleTitle=null, refAbstract=null), Reference(id=1215325304537662392, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=Xing Qianguo, Hu Chuanmin. Mapping macroalgal blooms in the Yellow Sea and East China Sea using HJ-1 and Landsat data: Application of a virtual baseline reflectance height technique[J]. Remote Sensing of Environment, 2016, 178: 113−126., articleTitle=null, refAbstract=null), Reference(id=1215325304625742779, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=薛瑞, 吴孟泉, 刘杨, 等. 基于HJ-1A/1B的2014年黄海海域浒苔灾害时空分布[J]. 海洋科学, 2016, 40(7): 115−123., articleTitle=null, refAbstract=null), Reference(id=1215325304726406077, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=Xue Rui, Wu Mengquan, Liu Yang, et al. Spatial and temporal variability of Ulva prolifera in the Yellow Sea, China in 2014[J]. Marine Sciences, 2016, 40(7): 115−123., articleTitle=null, refAbstract=null), Reference(id=1215325304827069376, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=曾韬, 刘建强. “北京一号”小卫星在青岛近海浒苔灾害监测中的应用[J]. 遥感信息, 2009, 24(3): 34−37., articleTitle=null, refAbstract=null), Reference(id=1215325306081166275, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=Zeng Tao, Liu Jianqiang. The application of Beijing-1 micro satellite data to algae disaster monitoring in the sea of Qingdao[J]. Remote Sensing Information, 2009, 24(3): 34−37., articleTitle=null, refAbstract=null), Reference(id=1215325306169246664, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=刘锦超, 刘建强, 丁静, 等. HY-1C卫星CZI载荷的黄海绿潮提取研究[J]. 海洋学报, 2022, 44(5): 1−11., articleTitle=null, refAbstract=null), Reference(id=1215325306269909963, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=22, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Jinchao, Liu Jianqiang, Ding Jing, et al. A refined imagery algorithm to extract green tide in the Yellow Sea from HY-1C satellite CZI measurements[J]. Haiyang Xuebao, 2022, 44(5): 1−11., articleTitle=null, refAbstract=null), Reference(id=1215325306357990351, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=23, authorNames=null, journalName=null, refType=null, unstructuredReference=He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, USA: IEEE, 2016: 770−778., articleTitle=null, refAbstract=null), Reference(id=1215325306433487827, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=Badrinarayanan V, Kendall A, Cipolla R. SegNet: a deep convolutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12): 2481−2495., articleTitle=null, refAbstract=null), Reference(id=1215325306534151126, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=25, authorNames=null, journalName=null, refType=null, unstructuredReference=Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation[C]//18th International Conference on Medical Image Computing and Computer-Assisted Intervention. Munich: Springer, 2015: 234−241., articleTitle=null, refAbstract=null), Reference(id=1215325306592871385, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=26, authorNames=null, journalName=null, refType=null, unstructuredReference=Cui Binge, Zhang Haoqing, Jing Wei, et al. SRSe-net: super-resolution-based semantic segmentation network for green tide extraction[J]. Remote Sensing, 2022, 14(3): 710., articleTitle=null, refAbstract=null), Reference(id=1215325306647397340, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=27, authorNames=null, journalName=null, refType=null, unstructuredReference=Yu Haifei, Wang Changying, Li Jinhua, et al. Automatic extraction of green tide from GF-3 SAR images based on feature selection and deep learning[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021, 14: 10598−10613., articleTitle=null, refAbstract=null), Reference(id=1215325306714506207, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=28, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Zhongyuan, Fang Zhixiang, Liang Jianfeng, et al. Estimating Ulva prolifera green tides of the Yellow Sea through ConvLSTM data fusion[J]. Environmental Pollution, 2023, 324: 121350., articleTitle=null, refAbstract=null), Reference(id=1215325306781615073, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=29, authorNames=null, journalName=null, refType=null, unstructuredReference=Shang Weitao, Gao Zhiqiang, Gao Meng, et al. Monitoring green tide in the Yellow Sea using high-resolution imagery and deep learning[J]. Remote Sensing, 2023, 15(4): 1101., articleTitle=null, refAbstract=null), Reference(id=1215325306857112549, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=26, rfOrder=30, authorNames=null, journalName=null, refType=null, unstructuredReference=Shi Wei, Wang Menghua. Green macroalgae blooms in the Yellow Sea during the spring and summer of 2008[J]. Journal of Geophysical Research: Oceans, 2009, 114(C12): C12010., articleTitle=null, refAbstract=null), Reference(id=1215325306953581545, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=27, rfOrder=31, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Mengqiu, Hu Chuanmin. Mapping and quantifying Sargassum distribution and coverage in the Central West Atlantic using MODIS observations[J]. Remote Sensing of Environment, 2016, 183: 350−367., articleTitle=null, refAbstract=null), Reference(id=1215325307037467628, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=28, rfOrder=32, authorNames=null, journalName=null, refType=null, unstructuredReference=苗珊珊. “长征”二号C运载火箭成功发射“海洋”一号C卫星[J]. 中国航天, 2018(9): 26., articleTitle=null, refAbstract=null), Reference(id=1215325307230405615, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=28, rfOrder=33, authorNames=null, journalName=null, refType=null, unstructuredReference=Miao Shanshan. The CZ-2C carrier rocket successfully launched the HY-1C satellite[J]. Aerospace China, 2018(9): 26., articleTitle=null, refAbstract=null), Reference(id=1215325307318486003, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=29, rfOrder=34, authorNames=null, journalName=null, refType=null, unstructuredReference=刘建强, 蒋兴伟, 王丽丽, 等. 海洋一号C、D卫星组网观测与应用[J]. 卫星应用, 2021(9): 19−26., articleTitle=null, refAbstract=null), Reference(id=1215325307373011959, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=29, rfOrder=35, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Jianqiang, Jiang Xingwei, Wang Lili, et al. Monitoring and application of the HY-1C and D satellites constellation[J]. Satellite Application, 2021(9): 19−26., articleTitle=null, refAbstract=null), Reference(id=1215325307452703740, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=30, rfOrder=36, authorNames=null, journalName=null, refType=null, unstructuredReference=李岩松, 赵慧洁, 李娜, 等. 基于中红外偏振的海面太阳耀光背景下的目标探测[J]. 中国激光, 2022, 49(19): 1910004., articleTitle=null, refAbstract=null), Reference(id=1215325307557561341, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=30, rfOrder=37, authorNames=null, journalName=null, refType=null, unstructuredReference=Li Yansong, Zhao Huijie, Li Na, et al. Detection of marine targets covered in sun glint based on mid-infrared polarization[J]. Chinese Journal of Lasers, 2022, 49(19): 1910004., articleTitle=null, refAbstract=null), Reference(id=1215325307658224640, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=31, rfOrder=38, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Chen, Zhang Huaguo, Xu Qing, et al. Inversion of the refractive index of marine spilled oil using multi-angle sun glitter images acquired by the ASTER sensor[J]. Remote Sensing of Environment, 2022, 275: 113019., articleTitle=null, refAbstract=null), Reference(id=1215325307737915392, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=32, rfOrder=39, authorNames=null, journalName=null, refType=null, unstructuredReference=刘建强, 曾韬, 梁超, 等. 海洋一号C卫星在自然灾害监测中的应用[J]. 卫星应用, 2020(6): 26−34., articleTitle=null, refAbstract=null), Reference(id=1215325307796635651, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=32, rfOrder=40, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Jianqiang, Zeng Tao, Liang Chao, et al. Application of HY-1C satellite in natural disaster monitoring[J]. Satellite Application, 2020(6): 26−34., articleTitle=null, refAbstract=null)], funds=[Fund(id=1215325302855746425, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, awardId=null, language=CN, fundingSource=国家自然科学基金项目(62172247);山东省重点研发计划重大科技创新工程项目(2019JZZY020101)。, fundOrder=null, country=null)], companyList=[AuthorCompany(id=1215325293666025990, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, xref=1, ext=[AuthorCompanyExt(id=1215325293674414599, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, companyId=1215325293666025990, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 青岛大学 计算机科学技术学院,山东 青岛, 266071)]), AuthorCompany(id=1215325293850575378, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, xref=1, ext=[AuthorCompanyExt(id=1215325293858963987, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, companyId=1215325293850575378, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1School of Computer Science and Technology, Qingdao University, Qingdao 266071, China)])], figs=[ArticleFig(id=1215325297562534576, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 1, caption=Data distribution of band 3 of green tide, sea water and thick cloud

Data is randomly sampled from the HY-1C/D satellite CZI sensor images on May 25, June 21, and June 30, 2001. The number of green tides and seawater samples is about 10 000. The band 3 data distribution of thick cloud sample points is too centralized to be easily displayed. The number of band 3 is about 5 000

, figureFileSmall=lMxXCHI3XmUOBIq0pEK1NQ==, figureFileBig=SQis5inbOqBcRMHKM3rvWQ==, tableContent=null), ArticleFig(id=1215325297621254835, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图1, caption=绿潮、海水、厚云band 3波段的数据分布

数据随机取样自2021年5月25日、6月21日、6月30日HY-1C/D卫星CZI传感器影像;绿潮和海水样本点数量约为10 000个;厚云样本点band 3数据分布过于集中,为便于展示,其数量约为5 000个

, figureFileSmall=lMxXCHI3XmUOBIq0pEK1NQ==, figureFileBig=SQis5inbOqBcRMHKM3rvWQ==, tableContent=null), ArticleFig(id=1215325297747083963, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 2, caption=Rule B extraction results (green), figureFileSmall=TTzzT02kh8h71i4/Cv/1qg==, figureFileBig=oAO1UE0gPmwWNGI3uxlCdA==, tableContent=null), ArticleFig(id=1215325297830970049, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图2, caption=规则B提取结果(绿色部分), figureFileSmall=TTzzT02kh8h71i4/Cv/1qg==, figureFileBig=oAO1UE0gPmwWNGI3uxlCdA==, tableContent=null), ArticleFig(id=1215325297935827654, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 5, caption=Rule E extraction results (orange)

There are no properly detected green tide cells in this area, so no artificial interpretation labels are given

, figureFileSmall=PjveGESbEXIyIhuPBj6rTg==, figureFileBig=Ah94j5Sdijiw0eLsrjzcqA==, tableContent=null), ArticleFig(id=1215325298040685259, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图5, caption=规则E提取结果(橙色部分)

该区域不存在正确检测的绿潮像元,故未给出人工解译标签

, figureFileSmall=PjveGESbEXIyIhuPBj6rTg==, figureFileBig=Ah94j5Sdijiw0eLsrjzcqA==, tableContent=null), ArticleFig(id=1215325298141348559, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 3, caption=Rule C extraction results (red), figureFileSmall=/jHBuT+033mf+ZLVypXsWw==, figureFileBig=5K2+RwjEK15sMFsGPo1n+w==, tableContent=null), ArticleFig(id=1215325298254594773, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图3, caption=规则C提取结果(红色部分), figureFileSmall=/jHBuT+033mf+ZLVypXsWw==, figureFileBig=5K2+RwjEK15sMFsGPo1n+w==, tableContent=null), ArticleFig(id=1215325298342675161, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 4, caption=Rule D extraction results (blue), figureFileSmall=OpHQzYX/kuZ4/Ogbhe+MpQ==, figureFileBig=P9UxyvG2AjK5gyQZv/kL2g==, tableContent=null), ArticleFig(id=1215325298464309985, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图4, caption=规则D提取结果(蓝色部分), figureFileSmall=OpHQzYX/kuZ4/Ogbhe+MpQ==, figureFileBig=P9UxyvG2AjK5gyQZv/kL2g==, tableContent=null), ArticleFig(id=1215325298552390374, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 6, caption=Error category correction process, figureFileSmall=IDKfcRxKvEWHMBHThTX2eg==, figureFileBig=Q5p70ArPXpZRDBpInS6wbg==, tableContent=null), ArticleFig(id=1215325298707579627, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图6, caption=错误类别修正流程, figureFileSmall=IDKfcRxKvEWHMBHThTX2eg==, figureFileBig=Q5p70ArPXpZRDBpInS6wbg==, tableContent=null), ArticleFig(id=1215325298825020143, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 7, caption=Comparison of classification results

a. HY-1C/D satellite CZI sensor RGB synthetic images (R: 825 nm, G: 650 nm, B: 560 nm); b. classification results based on rule sets; c. classification results after correcting error categories

, figureFileSmall=S3lAKNKDlO9GW8gc1fRQpA==, figureFileBig=s5uIP3i8m32lWSosNoomSg==, tableContent=null), ArticleFig(id=1215325298908906224, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图7, caption=分类结果对比

a. HY-1C/D卫星CZI传感器RGB合成影像(R:825 nm,G:650 nm,B:560 nm);b. 基于规则集的分类结果;c. 修正错误类别后的分类结果

, figureFileSmall=S3lAKNKDlO9GW8gc1fRQpA==, figureFileBig=s5uIP3i8m32lWSosNoomSg==, tableContent=null), ArticleFig(id=1215325299043123960, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 8, caption=Full-automatic extraction method of green tide from HY-1C/D CZI images, figureFileSmall=+Uc2sqCU3QhQZk7+n/9J/w==, figureFileBig=NTpVhqw2iStIETvC71E/dA==, tableContent=null), ArticleFig(id=1215325299122815743, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图8, caption=HY-1C/D CZI影像绿潮全自动提取方法, figureFileSmall=+Uc2sqCU3QhQZk7+n/9J/w==, figureFileBig=NTpVhqw2iStIETvC71E/dA==, tableContent=null), ArticleFig(id=1215325299257033475, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 9, caption=Scatter plot of NDVI and VB-FAH data

Both green tide and non green tide data are randomly sampled, with a quantity of 1 000

, figureFileSmall=VSwVdGq8x1pVFyIbK7xbWw==, figureFileBig=Yq1kX0jsW1Vc2Xr7AMlvRg==, tableContent=null), ArticleFig(id=1215325299395445512, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图9, caption=NDVI和VB-FAH数据散点图

绿潮与非绿潮数据均为随机取样,数量为1 000个

, figureFileSmall=VSwVdGq8x1pVFyIbK7xbWw==, figureFileBig=Yq1kX0jsW1Vc2Xr7AMlvRg==, tableContent=null), ArticleFig(id=1215325299538051857, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 10, caption=Accuracy evaluation area

a, b, c. HY-1C/D satellite CZI sensor RGB synthetic image (R: 825 nm, G: 650 nm, B: 560 nm); area 1–20 with dimensions of 400 pixels × 400 pixels; due to the fragmented green tide patches in the flare area, the size of the area 21–25 is set to 100 pixels × 100 pixels; each pixel size is 50 m × 50 m

, figureFileSmall=fzM1jBLbOtOgDyCItUW0Iw==, figureFileBig=wvyjvuMrM64QA42C24luNQ==, tableContent=null), ArticleFig(id=1215325299647103769, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图10, caption=精度评估区域

a、b、c. HY-1C/D卫星CZI传感器RGB合成影像(R:825 nm,G:650 nm,B:560 nm);区域1至区域20尺寸为400像素 × 400像素;由于耀斑区域中绿潮斑块较零碎,故将区域21至区域25尺寸设定为100像素 × 100像素;每个像素尺寸为50 m × 50 m

, figureFileSmall=fzM1jBLbOtOgDyCItUW0Iw==, figureFileBig=wvyjvuMrM64QA42C24luNQ==, tableContent=null), ArticleFig(id=1215325299747767071, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 11, caption=Comparison of green tide extraction effects in thick cloud region

a. RGB composite image (R: 825 nm, G: 650 nm, B: 560 nm) of HY-1C/D satellite CZI sensor; b, c, d, e, f, g show the green tide extraction results of manual interpretation, NDVI, VB-FAH, ResNet50, U-Net, and the method proposed in this paper, respectively

, figureFileSmall=yFgv45BBH4OM/ZHZsS0uxw==, figureFileBig=yVQNIOyN9yXAyDI6NnhqUQ==, tableContent=null), ArticleFig(id=1215325299869401892, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图11, caption=厚云区域绿潮提取效果对比

a. HY-1C/D卫星CZI传感器RGB合成影像(R:825 nm,G:650 nm,B:560 nm);b、c、d、e、f、g分别为人工解译、NDVI、VB-FAH、ResNet50、U-Net、本文方法的绿潮提取结果

, figureFileSmall=yFgv45BBH4OM/ZHZsS0uxw==, figureFileBig=yVQNIOyN9yXAyDI6NnhqUQ==, tableContent=null), ArticleFig(id=1215325300024591149, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 12, caption=Comparison of green tide extraction effects in thin cloud region

a. RGB composite image (R: 825 nm, G: 650 nm, B: 560 nm) of HY-1C/D satellite CZI sensor; b, c, d, e, f, g show the green tide extraction results of manual interpretation, NDVI, VB-FAH, ResNet50, U-Net, and the method proposed in this paper, respectively

, figureFileSmall=6Zd/BQkSKlMKMhSR9zNrPA==, figureFileBig=Or2ykKXjrIhz7iqpZSSFfQ==, tableContent=null), ArticleFig(id=1215325300121060147, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图12, caption=薄云区域绿潮提取效果对比

a. HY-1C/D卫星CZI传感器RGB合成影像(R:825 nm,G:650 nm,B:560 nm);b、c、d、e、f、g分别为人工解译、NDVI、VB-FAH、ResNet50、U-Net、本文方法的绿潮提取结果

, figureFileSmall=6Zd/BQkSKlMKMhSR9zNrPA==, figureFileBig=Or2ykKXjrIhz7iqpZSSFfQ==, tableContent=null), ArticleFig(id=1215325300209140534, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 13, caption=Comparison of green tide extraction effects in cloud-free region

a. RGB composite image (R: 825 nm, G: 650 nm, B: 560 nm) of HY-1C/D satellite CZI sensor; b, c, d, e, f, g show the green tide extraction results of manual interpretation, NDVI, VB-FAH, ResNet50, U-Net, and the method proposed in this paper, respectively

, figureFileSmall=AZBQTmKc8Cg8hCfPs2u80w==, figureFileBig=iAbysXOYm3pTJhC0FP1t5w==, tableContent=null), ArticleFig(id=1215325300288832314, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图13, caption=无云区域绿潮提取效果对比

a. HY-1C/D卫星CZI传感器RGB合成影像(R:825 nm,G:650 nm,B:560 nm);b、c、d、e、f、g分别为人工解译、NDVI、VB-FAH、ResNet50、U-Net、本文方法的绿潮提取结果

, figureFileSmall=AZBQTmKc8Cg8hCfPs2u80w==, figureFileBig=iAbysXOYm3pTJhC0FP1t5w==, tableContent=null), ArticleFig(id=1215325300351746877, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 14, caption=Comparison of green tide extraction effects in cloud spot region

a. RGB composite image (R: 825 nm, G: 650 nm, B: 560 nm) of HY-1C/D satellite CZI sensor; b, c, d, e, f, g show the green tide extraction results of manual interpretation, NDVI, VB-FAH, ResNet50, U-Net, and the method proposed in this paper, respectively

, figureFileSmall=LEL/CSBORdLDRtnZITEFxA==, figureFileBig=3JsZAKMHA1r8TS4PAnf/RQ==, tableContent=null), ArticleFig(id=1215325301601649473, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图14, caption=云斑区域绿潮提取效果对比

a. HY-1C/D卫星CZI传感器RGB合成影像(R:825 nm,G:650 nm,B:560 nm);b、c、d、e、f、g分别为人工解译、NDVI、VB-FAH、ResNet50、U-Net、本文方法的绿潮提取结果

, figureFileSmall=LEL/CSBORdLDRtnZITEFxA==, figureFileBig=3JsZAKMHA1r8TS4PAnf/RQ==, tableContent=null), ArticleFig(id=1215325301681341252, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Fig. 15, caption=Comparison of green tide extraction effect in flare region

a. RGB composite image (R: 825 nm, G: 650 nm, B: 560 nm) of HY-1C/D satellite CZI sensor; b, c, d, e, f, g show the green tide extraction results of manual interpretation, NDVI, VB-FAH, ResNet50, U-Net, and the method proposed in this paper, respectively

, figureFileSmall=T9CeOk4hHTKVMHYgbnLtCA==, figureFileBig=GyFPNCG0Ij4nGtc3NvT6Wg==, tableContent=null), ArticleFig(id=1215325301765227335, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=图15, caption=耀斑区域绿潮提取效果对比

a. HY-1C/D卫星CZI传感器RGB合成影像(R:825 nm,G:650 nm,B:560 nm);b、c、d、e、f、g分别为人工解译、NDVI、VB-FAH、ResNet50、U-Net、本文方法的绿潮提取结果

, figureFileSmall=T9CeOk4hHTKVMHYgbnLtCA==, figureFileBig=GyFPNCG0Ij4nGtc3NvT6Wg==, tableContent=null), ArticleFig(id=1215325301849113420, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Table 1, caption=

Band information of HY-1C/D satellite CZI sensor

, figureFileSmall=null, figureFileBig=null, tableContent=
波段波宽/nm空间分辨率/m
band 1420~50050
band 2520~60050
band 3610~69050
band 4760~89050
), ArticleFig(id=1215325301958165330, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=表1, caption=

HY-1C/D 卫星 CZI 传感器的波段信息

, figureFileSmall=null, figureFileBig=null, tableContent=
波段波宽/nm空间分辨率/m
band 1420~50050
band 2520~60050
band 3610~69050
band 4760~89050
), ArticleFig(id=1215325302067217236, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Table 2, caption=

Green tide extraction rule set

, figureFileSmall=null, figureFileBig=null, tableContent=
编号决策规则颜色
Ab3 – b4 ≤ –389.5且b3 – b4 ≤ –328.5且b1 ≤ 448.5若不满足A,则颜色为黑色
b3 – b4 > –389.5且b2 – b3 > –84.5且b3 – b4 >863
b3 – b4 > –389.5且b2 – b3 > –84.5且b3 – b4 > –523.5
b3 – b4 > –389.5且b2 – b3 > –84.5且b3 – b4 ≤ –523.5且b3 ≤ –484.5
BA且b2 – b3 ≤ –140.5 且 b1 ≤ 1558且b2 – b3 > –4.5绿色
A且b2 – b3 ≤ –140.5且b2 – b3 > –321.5
A 且b2 – b3 ≤ –140.5且b1 ≤ 1558且b2 – b3 ≤ –4.5且b3 – b4 ≤ –681.5且
b2 – b3 > –42.5
CA 且 b2 – b3 ≤ –140.5且b1 > 1558红色
DA且b2 – b3 ≤ –140.5且b1 ≤ 1558且b2 – b3 ≤ –4.5且b3 – b4 ≤ –681.5且
b2 – b3 ≤ –42.5
蓝色
A且b2 – b3 ≤ –140.5且b1 ≤ 1558且b2 – b3 ≤ –4.5且b3 – b4 > –681.5
EA且b2 – b3 ≤ –140.5且b2 – b3 ≤ –321.5橙色
), ArticleFig(id=1215325302172074841, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=表2, caption=

绿潮提取规则集

, figureFileSmall=null, figureFileBig=null, tableContent=
编号决策规则颜色
Ab3 – b4 ≤ –389.5且b3 – b4 ≤ –328.5且b1 ≤ 448.5若不满足A,则颜色为黑色
b3 – b4 > –389.5且b2 – b3 > –84.5且b3 – b4 >863
b3 – b4 > –389.5且b2 – b3 > –84.5且b3 – b4 > –523.5
b3 – b4 > –389.5且b2 – b3 > –84.5且b3 – b4 ≤ –523.5且b3 ≤ –484.5
BA且b2 – b3 ≤ –140.5 且 b1 ≤ 1558且b2 – b3 > –4.5绿色
A且b2 – b3 ≤ –140.5且b2 – b3 > –321.5
A 且b2 – b3 ≤ –140.5且b1 ≤ 1558且b2 – b3 ≤ –4.5且b3 – b4 ≤ –681.5且
b2 – b3 > –42.5
CA 且 b2 – b3 ≤ –140.5且b1 > 1558红色
DA且b2 – b3 ≤ –140.5且b1 ≤ 1558且b2 – b3 ≤ –4.5且b3 – b4 ≤ –681.5且
b2 – b3 ≤ –42.5
蓝色
A且b2 – b3 ≤ –140.5且b1 ≤ 1558且b2 – b3 ≤ –4.5且b3 – b4 > –681.5
EA且b2 – b3 ≤ –140.5且b2 – b3 ≤ –321.5橙色
), ArticleFig(id=1215325302260155228, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Table 3, caption=

Accuracy evaluation results of ACC and Kappa

, figureFileSmall=null, figureFileBig=null, tableContent=
区域/类型ACCKappa
NDVIVB-FAHResNet50U-Net本文方法NDVIVB-FAHResNet50U-Net本文方法
1/厚云0.959 20.912 70.930 10.959 50.992 30.814 80.508 00.614 40.791 90.964 5
2/厚云0.803 70.791 40.980 70.988 20.994 10.156 90.065 40.589 80.761 00.891 6
3/厚云0.901 20.875 10.962 40.976 10.993 20.400 50.096 50.550 30.748 20.935 2
4/厚云0.927 40.929 40.984 30.990 20.992 50.462 00.381 00.766 30.600 30.894 1
5/厚云0.872 20.869 00.976 10.982 60.990 10.389 90.334 30.771 90.643 40.842 4
6/薄云0.969 40.947 20.957 90.970 00.970 90.824 50.583 20.710 70.694 50.849 2
7/薄云0.966 70.968 30.976 10.982 00.989 40.719 10.558 40.744 20.698 00.885 4
8/薄云0.977 50.973 70.976 40.983 90.985 50.815 20.691 10.767 70.837 80.859 6
9/薄云0.965 10.984 60.984 80.970 10.987 10.775 50.864 30.879 00.922 50.982 5
10/薄云0.948 70.927 10.924 60.945 20.994 50.774 70.523 50.587 60.691 30.973 1
11/无云0.904 90.980 90.978 50.969 10.990 40.576 80.853 60.852 60.825 00.921 3
12/无云0.949 20.971 50.966 70.967 20.979 60.673 10.688 70.707 20.783 30.828 2
13/无云0.923 40.957 40.963 10.968 50.970 40.682 40.726 90.801 00.678 20.847 3
14/无云0.975 00.975 50.973 50.985 70.998 20.769 70.621 30.683 50.820 60.978 5
15/无云0.985 00.977 40.980 80.986 90.989 00.830 70.627 00.732 60.815 90.850 1
16/云斑0.905 60.955 60.940 40.971 80.982 30.606 80.660 40.612 70.832 90.896 5
17/云斑0.914 10.969 80.962 20.974 20.994 10.470 00.483 80.500 90.676 10.927 7
18/云斑0.868 00.957 50.944 90.955 70.985 10.454 30.549 80.572 50.698 50.887 9
19/云斑0.924 00.981 10.976 10.977 70.984 00.478 80.682 70.683 80.856 70.906 0
20/云斑0.904 20.955 40.944 70.967 80.984 10.603 10.661 50.682 50.817 40.907 8
21/耀斑0.907 50.978 10.977 50.976 40.995 60.334 60.294 50.252 30.187 70.919 1
22/耀斑0.896 60.987 50.988 60.988 60.995 10.188 50.147 90.275 70.275 70.819 0
23/耀斑0.903 90.987 60.987 30.987 30.996 70.219 40.664 60.279 40.279 40.890 9
24/耀斑0.959 30.993 10.994 60.994 60.997 20.254 60.125 60.423 70.423 70.816 0
25/耀斑0.711 90.966 10.981 50.981 50.987 10.167 80.353 70.730 70.730 70.860 4
), ArticleFig(id=1215325302385984353, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=表3, caption=

ACC和Kappa精度评估结果

, figureFileSmall=null, figureFileBig=null, tableContent=
区域/类型ACCKappa
NDVIVB-FAHResNet50U-Net本文方法NDVIVB-FAHResNet50U-Net本文方法
1/厚云0.959 20.912 70.930 10.959 50.992 30.814 80.508 00.614 40.791 90.964 5
2/厚云0.803 70.791 40.980 70.988 20.994 10.156 90.065 40.589 80.761 00.891 6
3/厚云0.901 20.875 10.962 40.976 10.993 20.400 50.096 50.550 30.748 20.935 2
4/厚云0.927 40.929 40.984 30.990 20.992 50.462 00.381 00.766 30.600 30.894 1
5/厚云0.872 20.869 00.976 10.982 60.990 10.389 90.334 30.771 90.643 40.842 4
6/薄云0.969 40.947 20.957 90.970 00.970 90.824 50.583 20.710 70.694 50.849 2
7/薄云0.966 70.968 30.976 10.982 00.989 40.719 10.558 40.744 20.698 00.885 4
8/薄云0.977 50.973 70.976 40.983 90.985 50.815 20.691 10.767 70.837 80.859 6
9/薄云0.965 10.984 60.984 80.970 10.987 10.775 50.864 30.879 00.922 50.982 5
10/薄云0.948 70.927 10.924 60.945 20.994 50.774 70.523 50.587 60.691 30.973 1
11/无云0.904 90.980 90.978 50.969 10.990 40.576 80.853 60.852 60.825 00.921 3
12/无云0.949 20.971 50.966 70.967 20.979 60.673 10.688 70.707 20.783 30.828 2
13/无云0.923 40.957 40.963 10.968 50.970 40.682 40.726 90.801 00.678 20.847 3
14/无云0.975 00.975 50.973 50.985 70.998 20.769 70.621 30.683 50.820 60.978 5
15/无云0.985 00.977 40.980 80.986 90.989 00.830 70.627 00.732 60.815 90.850 1
16/云斑0.905 60.955 60.940 40.971 80.982 30.606 80.660 40.612 70.832 90.896 5
17/云斑0.914 10.969 80.962 20.974 20.994 10.470 00.483 80.500 90.676 10.927 7
18/云斑0.868 00.957 50.944 90.955 70.985 10.454 30.549 80.572 50.698 50.887 9
19/云斑0.924 00.981 10.976 10.977 70.984 00.478 80.682 70.683 80.856 70.906 0
20/云斑0.904 20.955 40.944 70.967 80.984 10.603 10.661 50.682 50.817 40.907 8
21/耀斑0.907 50.978 10.977 50.976 40.995 60.334 60.294 50.252 30.187 70.919 1
22/耀斑0.896 60.987 50.988 60.988 60.995 10.188 50.147 90.275 70.275 70.819 0
23/耀斑0.903 90.987 60.987 30.987 30.996 70.219 40.664 60.279 40.279 40.890 9
24/耀斑0.959 30.993 10.994 60.994 60.997 20.254 60.125 60.423 70.423 70.816 0
25/耀斑0.711 90.966 10.981 50.981 50.987 10.167 80.353 70.730 70.730 70.860 4
), ArticleFig(id=1215325302516007784, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=EN, label=Table 4, caption=

Accuracy evaluation results of F1-Score and MIoU

, figureFileSmall=null, figureFileBig=null, tableContent=
区域/类型F1-ScoreMIoU
NDVIVB-FAHResNet50U-Net本文方法NDVIVB-FAHResNet50U-Net本文方法
1/厚云0.967 00.945 50.958 20.977 00.993 70.837 90.644 60.703 70.821 00.965 5
2/厚云0.803 40.796 60.988 30.993 40.996 40.455 30.424 60.704 00.802 40.901 6
3/厚云0.907 40.885 70.978 30.985 70.995 60.592 60.476 10.679 00.794 40.938 7
4/厚云0.926 50.933 80.988 50.990 10.993 70.625 00.593 60.807 80.908 70.963 6
5/厚云0.874 00.875 40.975 60.968 60.982 20.576 00.554 10.810 10.862 40.961 3
6/薄云0.974 70.972 90.975 30.964 60.983 20.846 50.690 30.766 70.824 50.935 3
7/薄云0.969 00.983 80.983 60.969 50.987 50.773 90.684 00.791 90.829 10.920 1
8/薄云0.978 80.986 60.984 10.970 60.990 30.840 60.757 80.807 40.858 30.875 2
9/薄云0.9640.931 90.928 20.961 60.9880.810 30.878 70.890 50.927 40.950 1
10/薄云0.952 60.962 10.949 90.968 20.997 10.807 40.653 80.688 70.751 90.973 7
11/无云0.901 10.970 10.973 40.951 50.980 70.674 50.869 90.869 00.929 50.944 0
12/无云0.949 20.975 40.975 10.961 90.983 50.741 50.755 80.766 90.894 20.950 6
13/无云0.921 80.978 10.972 50.967 50.985 20.741 10.776 40.828 40.889 20.963 9
14/无云0.974 80.987 60.980 70.971 10.999 00.808 20.718 60.754 00.845 80.978 9
15/无云0.986 60.988 60.988 00.992 80.994 00.853 20.722 20.785 30.842 60.868 3
16/云斑0.901 80.976 90.957 30.976 90.984 40.691 40.735 40.706 50.853 10.904 6
17/云斑0.912 60.984 70.974 10.982 10.996 50.624 10.649 30.656 40.749 80.932 2
18/云斑0.863 60.978 10.957 90.960 80.987 60.600 60.676 50.686 60.759 10.897 8
19/云斑0.922 60.950 20.982 00.968 00.986 10.631 40.754 60.754 70.773 10.895 1
20/云斑0.900 40.976 50.954 40.971 80.985 70.689 00.736 00.747 30.840 90.914 3
21/耀斑0.906 20.988 80.988 60.988 10.996 30.563 50.577 30.562 60.541 20.924 8
22/耀斑0.895 90.993 70.994 30.994 30.997 50.505 70.534 20.575 20.575 20.818 1
23/耀斑0.903 20.993 80.993 60.993 60.997 60.519 50.745 70.575 90.575 90.901 3
24/耀斑0.959 20.996 50.997 20.997 20.998 40.555 50.530 30.632 40.632 40.823 3
25/耀斑0.705 30.982 80.974 40.957 80.987 60.415 00.594 10.784 00.784 00.876 0
), ArticleFig(id=1215325302650225518, tenantId=1146029695717560320, journalId=1149651085930835976, articleId=1212062584971334532, language=CN, label=表4, caption=

F1-Score和MIoU精度评估结果

, figureFileSmall=null, figureFileBig=null, tableContent=
区域/类型F1-ScoreMIoU
NDVIVB-FAHResNet50U-Net本文方法NDVIVB-FAHResNet50U-Net本文方法
1/厚云0.967 00.945 50.958 20.977 00.993 70.837 90.644 60.703 70.821 00.965 5
2/厚云0.803 40.796 60.988 30.993 40.996 40.455 30.424 60.704 00.802 40.901 6
3/厚云0.907 40.885 70.978 30.985 70.995 60.592 60.476 10.679 00.794 40.938 7
4/厚云0.926 50.933 80.988 50.990 10.993 70.625 00.593 60.807 80.908 70.963 6
5/厚云0.874 00.875 40.975 60.968 60.982 20.576 00.554 10.810 10.862 40.961 3
6/薄云0.974 70.972 90.975 30.964 60.983 20.846 50.690 30.766 70.824 50.935 3
7/薄云0.969 00.983 80.983 60.969 50.987 50.773 90.684 00.791 90.829 10.920 1
8/薄云0.978 80.986 60.984 10.970 60.990 30.840 60.757 80.807 40.858 30.875 2
9/薄云0.9640.931 90.928 20.961 60.9880.810 30.878 70.890 50.927 40.950 1
10/薄云0.952 60.962 10.949 90.968 20.997 10.807 40.653 80.688 70.751 90.973 7
11/无云0.901 10.970 10.973 40.951 50.980 70.674 50.869 90.869 00.929 50.944 0
12/无云0.949 20.975 40.975 10.961 90.983 50.741 50.755 80.766 90.894 20.950 6
13/无云0.921 80.978 10.972 50.967 50.985 20.741 10.776 40.828 40.889 20.963 9
14/无云0.974 80.987 60.980 70.971 10.999 00.808 20.718 60.754 00.845 80.978 9
15/无云0.986 60.988 60.988 00.992 80.994 00.853 20.722 20.785 30.842 60.868 3
16/云斑0.901 80.976 90.957 30.976 90.984 40.691 40.735 40.706 50.853 10.904 6
17/云斑0.912 60.984 70.974 10.982 10.996 50.624 10.649 30.656 40.749 80.932 2
18/云斑0.863 60.978 10.957 90.960 80.987 60.600 60.676 50.686 60.759 10.897 8
19/云斑0.922 60.950 20.982 00.968 00.986 10.631 40.754 60.754 70.773 10.895 1
20/云斑0.900 40.976 50.954 40.971 80.985 70.689 00.736 00.747 30.840 90.914 3
21/耀斑0.906 20.988 80.988 60.988 10.996 30.563 50.577 30.562 60.541 20.924 8
22/耀斑0.895 90.993 70.994 30.994 30.997 50.505 70.534 20.575 20.575 20.818 1
23/耀斑0.903 20.993 80.993 60.993 60.997 60.519 50.745 70.575 90.575 90.901 3
24/耀斑0.959 20.996 50.997 20.997 20.998 40.555 50.530 30.632 40.632 40.823 3
25/耀斑0.705 30.982 80.974 40.957 80.987 60.415 00.594 10.784 00.784 00.876 0
)], attaches=null, journal=Journal(id=1146441459026210850, delFlag=0, nameCn=海洋学报, nameEn=Haiyang Xuebao, nameHistory1=null, nameHistory2=null, issn=0253-4193, eissn=null, cn=11-2055/P, coden=null, periodic=0, language=CN, oaType=否, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=uYi7hkkrve+l8pIcwqcaQQ==, journalPrice=null, startedYear=null, abbrevIsoEn=null, journalRemark=null, publicationField=null, createdTime=1751262543687, updatedTime=1761729782936, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=H, firstLetterEn=H, subjectCode=Natural Sciences, subjectName=Natural Sciences, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=uYi7hkkrve+l8pIcwqcaQQ==, picEn=C0WLQb7uW3ok8EkkVOAGuw==, jcr=null, cjcr=null, exts=[JournalExt(id=1190344242636624294, language=CN, name=海洋学报, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.hyxbocean.cn/, createdTime=1761729782971, updatedTime=1761729782971, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.hyxb.org.cn/aos/ch/author/login.aspx, submissionEditorUrl=http://www.hyxb.org.cn/aos/ch/login.aspx, submissionReviewUrl=http://www.hyxb.org.cn/aos/ch/auditor/login.aspx, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1190344242712121767, language=EN, name=Haiyang Xuebao, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.aosocean.com/, createdTime=1761729782989, updatedTime=1761729782989, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.hyxb.org.cn/aos/ch/author/login.aspx, submissionEditorUrl=http://www.hyxb.org.cn/aos/ch/login.aspx, submissionReviewUrl=http://www.hyxb.org.cn/aos/ch/auditor/login.aspx, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1149651085930835976, websiteList=[Website(id=1188165202219512001, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1149651085930835976, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/hyxb/CN, language=CN, createTime=1761210259251, createBy=18614031015, updateTime=1761210330879, updateBy=18614031015, name=海洋学报-中文, tplId=1146099689490845704, title=海洋学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1188166688563413602, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202219512001, code=articleTextType, value=kx, createTime=1761210613623, updateTime=1761210613623, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166688538247775, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202219512001, code=banner, value=null, createTime=1761210613617, updateTime=1761210613617, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166688529859166, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202219512001, code=logo, value=https://castjournals.cast.org.cn/joweb/hyxb/CN/file/pic?fileId=BDEio/cxHnid8OD4QxrAYQ==, createTime=1761210613615, updateTime=1761210613615, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166688555024993, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202219512001, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/hyxb/CN/file/pic, createTime=1761210613621, updateTime=1761210613621, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166688546636384, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202219512001, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1761210613619, updateTime=1761210613619, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166688575996515, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202219512001, code=themeColor, value=null, createTime=1761210613626, updateTime=1761210613626, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166688596968036, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202219512001, code=themeStyle, value=null, createTime=1761210613631, updateTime=1761210613631, creator=18614031015, updator=18614031015)]), Website(id=1188165202282426564, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1149651085930835976, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/hyxb/EN, language=EN, createTime=1761210259266, createBy=18614031015, updateTime=1761210377920, updateBy=18614031015, name=海洋学报-英文, tplId=1146101810881728533, title=Haiyang Xuebao, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1188166798101856873, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202282426564, code=articleTextType, value=kx, createTime=1761210639739, updateTime=1761210639739, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166798076691046, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202282426564, code=banner, value=null, createTime=1761210639733, updateTime=1761210639733, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166798068302437, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202282426564, code=logo, value=https://castjournals.cast.org.cn/joweb/hyxb/EN/file/pic?fileId=BDEio/cxHnid8OD4QxrAYQ==, createTime=1761210639731, updateTime=1761210639731, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166798093468264, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202282426564, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/hyxb/EN/file/pic, createTime=1761210639737, updateTime=1761210639737, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166798085079655, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202282426564, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1761210639735, updateTime=1761210639735, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166798106051178, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202282426564, code=themeColor, value=null, createTime=1761210639740, updateTime=1761210639740, creator=18614031015, updator=18614031015), WebsiteProps(id=1188166798110245483, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1188165202282426564, code=themeStyle, value=null, createTime=1761210639741, updateTime=1761210639741, creator=18614031015, updator=18614031015)])], journalTitle=海洋学报, weixinUrl=null, journalUrl=http://www.hyxbocean.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Haiyang Xuebao, journalPhotoCn=uYi7hkkrve+l8pIcwqcaQQ==, journalPhotoEn=C0WLQb7uW3ok8EkkVOAGuw==, journalFirstLetter=H, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/hyxb/CN/10.12284/hyxb2023151, detailUrlEn=https://castjournals.cast.org.cn/joweb/hyxb/EN/10.12284/hyxb2023151, pdfUrlCn=https://castjournals.cast.org.cn/joweb/hyxb/CN/PDF/10.12284/hyxb2023151, pdfUrlEn=https://castjournals.cast.org.cn/joweb/hyxb/EN/PDF/10.12284/hyxb2023151, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
HY-1C/D CZI多光谱影像云覆盖与耀斑区域绿潮自动提取方法研究
收藏切换
PDF下载
吴克 1 , 王常颖 1, * , 黄睿 1 , 李华伟 1
海洋学报 | 论文 2023,45(10): 168-182
收起
收藏切换
海洋学报 | 论文 2023, 45(10): 168-182
HY-1C/D CZI多光谱影像云覆盖与耀斑区域绿潮自动提取方法研究
全屏
吴克1 , 王常颖1, * , 黄睿1, 李华伟1
作者信息
  • 1 青岛大学 计算机科学技术学院,山东 青岛, 266071
  • 吴克(1999—),男,河南省濮阳市人,研究方向为遥感大数据。E-mail:

通讯作者:

*王常颖(1980—),副教授,主要从事海洋复杂性与数据挖掘研究。E-mail:
Automatic extraction of green tide in areas with clouds or solar flares in HY-1C/D CZI multispectral images
Ke Wu1 , Changying Wang1, * , Rui Huang1, Huawei Li1
Affiliations
  • 1School of Computer Science and Technology, Qingdao University, Qingdao 266071, China
出版时间: 2023-10-01 doi: 10.12284/hyxb2023151
文章导航
收藏切换

针对多光谱影像受云、雾、太阳耀斑等因素的影响,难以实现高精度的绿潮自动提取的问题,本文以我国的HY-1C/D卫星CZI载荷多光谱影像为数据源,采用数据挖掘技术,通过探索绿潮区域与非绿潮区域的光谱分布差异,提出一种适用于HY-1C/D CZI影像的高精度、全自动绿潮提取方法。首先,分析有云区域和无云区域样本的光谱差异,给出厚云去除规则;其次,选取绿潮和非绿潮区域的样本,采用决策树算法生成绿潮提取规则;然后,针对薄云和厚云边界区域常常会出现误检绿潮的问题,设计了5种错误类别修正策略。为验证方法的有效性,收集2021年黄海区域绿潮暴发周期内的25景HY-1C/D CZI影像,开展绿潮自动检测实验。结果表明,与传统的NDVI方法、VB-FAH方法等指数方法以及ResNet50、U-Net等深度学习方法相比,本文方法在准确度、Kappa系数、F1-Score和MIoU等指标上均优于其他方法,而且能够实现在厚云、薄云、无云、云斑和耀斑区域复杂情况下的绿潮的高精度自动提取。

HY-1C/D卫星  /  绿潮提取  /  决策树  /  耀斑  /  云覆盖

Multispectral images are greatly affected by factors such as clouds, fog, and solar flares, which makes it difficult to automatically extract high-precision green tides under complex weather conditions. Based on the multi-spectral images of my country’s HY-1C/D satellite CZI payload, using data mining technology to explore the difference in data distribution between green tide areas and non-green tide areas, we propose a high-precision and fully automatic green tide extraction method , which can be applied to HY-1C/D CZI sensor data. First of all, the thick cloud area is removed by preliminary extraction rules to achieve preliminary classification. Then, the correctly classified green tide samples and non-green tide samples were used as positive and negative samples respectively, and these samples were used as experimental data to train the decision tree model, and the automatic extraction rules of green tide were obtained according to the model. Finally, 5 strategies for correcting misclassifications were designed to achieve fully automatic extraction of green tides. In order to verify the effectiveness of the method, we collected 25 images of the green tide outbreak period in the Yellow Sea in 2021 for automatic detection experiments, and compared the experimental results with traditional index methods (NDVI, VB-FAH) and deep learning methods (ResNet50, U-Net). The results showed that the method outperformed other methods in terms of accuracy, Kappa coefficient, F1-Score, and MIoU. The accuracy of green tide extraction was higher in areas with thick clouds, thin clouds, cloudless clouds, cloud spots, and flares.

HY-1C/D satellite  /  green tide extraction  /  decision tree  /  solar flare  /  cloud cover
吴克, 王常颖, 黄睿, 李华伟. HY-1C/D CZI多光谱影像云覆盖与耀斑区域绿潮自动提取方法研究. 海洋学报, 2023 , 45 (10) : 168 -182 . DOI: 10.12284/hyxb2023151
Ke Wu, Changying Wang, Rui Huang, Huawei Li. Automatic extraction of green tide in areas with clouds or solar flares in HY-1C/D CZI multispectral images[J]. Haiyang Xuebao, 2023 , 45 (10) : 168 -182 . DOI: 10.12284/hyxb2023151
绿潮是海水中某些大型绿藻(如浒苔)暴发性增殖或高度聚集所形成的海洋生态现象[12]。由于全球气候变化和水体富营养化的影响,沿海绿潮灾害变得越来越频繁[37],已经成为全球性的海洋环境灾害。卫星遥感技术因其观测范围大、具有高分辨率和方便快捷等优势,成为国内外研究绿潮灾害监测的重要技术。利用卫星遥感技术可以快速确定研究区域内某一时刻绿潮聚集情况,并实现持续检测、分析绿潮的漂移轨迹,为绿潮灾害研究和治理提供必要技术和数据支持。在我国,绿潮暴发的主要时段为每年的5月至8月,其中盛行期多在6月中旬至7月中旬。由于绿潮分布范围大、变化快和形状不一,传统的实地观测存在局限性[89]。因此,卫星遥感技术成为研究绿潮灾害的重要手段[10]
绿潮提取技术在国内外学者中得到了广泛的研究。Hu和He[11]通过RGB图像与归一化植被指数(NDVI)相结合的方法,分析了黄海海域绿潮分布情况和漂移路径。施英妮等[12]提出了归一化藻类指数(NDAI)法,其研究结果表明该算法的绿潮提取精度优于传统算法。Hu[13]提出了浮游藻类指数(FAI)法,表现出更强的稳定性。Son等[14]提出了绿藻指数(IGAG)法。Xing和Hu[15]提出了虚拟基线高度浮藻指数绿潮提取方法(VB-FAH)。薛瑞等[16]应用神经网络监督分类法对黄海绿潮分布情况进行了动态监测。曾韬和刘建强[17]通过分析绿潮光谱特性,并采用人机交互的方法对黄海绿潮进行提取。刘锦超等[18]提出了一种结合藻类缩放指数SAI与VB-FAH结合的绿潮提取方法。近年来,卷积神经网络(CNN),如残差神经网络ResNet[19]、用于语义像素级分割的深度全卷积神经网络结构SegNet[20]、用于生物医学图像分割的卷积网络U-Net[21]等被广泛应用于遥感影像地物分类任务中。例如,Cui等[22]提出了一种适用于大规模绿潮信息提取的深层语义分割网络(SRSe-Net),可以有效获取绿潮边界的详细信息。Yu等[23]设计了一个轻量级的语义分割网络Mobile-SegNet,可有效地降低绿潮提取所需的特征维数,提高绿潮检测精度和效率。Wang等[24]通过卷积LSTM方法提出了一种新的绿潮估计框架(GTEE)。Shang[25]设计了一种深度学习模型VGGUnet用于提取海滩与近岸绿潮特征并计算绿潮覆盖面积与生物量密度,与实地勘测结果基本一致。这些研究表明,绿潮提取技术正在不断发展,将为绿潮监测和管理提供更加精准、高效的手段。
针对遥感影像绿潮检测中像元分割阈值的选取问题,低分辨率下的“含藻”像元提取误检测程度往往更大[26],尤其在绿潮分布面积的计算方面[27]。近年来,我国自主研制的海洋水色业务卫星海洋一号C(HaiYang-1C,HY-1C)卫星和海洋一号D(HaiYang-1D,HY-1D)卫星的发射[28],搭载的海岸带成像仪(CZI)能够提供50 m空间分辨率、950 km刈幅的多光谱数据[29],为绿潮提取工作带来更高精度数据,但同时也带来了一定挑战。云斑、船尾迹、太阳耀斑[3031]等会影响绿潮的提取,对50 m分辨率的绿潮遥感检测带来不便。
本文采用决策树方法,成功发现了黄海海域绿潮自动提取规则,并根据各错误类别分布特点完成修正,提出了基于HY-1C/D CZI影像的绿潮全自动提取方法。文章介绍了实验数据的波段信息、研究区域概况、数据预处理流程和绿潮自动提取规则的实现过程。设计了5种错误类别修正策略,解决了绿潮自动提取规则实际运用中误检测像元问题,提高了提取精度。将该方法应用于不同条件下的影像,并与传统的方法和深度学习方法进行比较,证明了该方法的优势。最后总结本文工作,并展望下一步的研究工作。
HY-1C/D卫星是我国国产的海洋卫星,以海洋探测为主,兼顾陆地探测。主要应用于全球海洋水色要素探测、海岸带动态环境监测和海表温度探测等任务。配备有海洋水色水温扫描仪(COCTS)、海岸带成像仪(CZI)、紫外成像仪(UVI)、船舶自动识别系统(AIS)和定标光谱仪等载荷[32]。其中,多光谱海岸带成像仪是本文研究所使用的数据源。
光谱海岸带成像仪包含可见光和近红外4个光谱波段,影像幅宽≥ 950 km,2~3景影像即可覆盖本文研究区域,像素分辨率为50 m,相较于低分辨率的MODIS数据,更有利于观测海上地物。因此,利用HY-1C/D卫星CZI数据进行绿潮监测研究具有显著优势。CZI数据的波段信息如表1所示。
本研究采用2021年5−8月间黄海区域的HY-1C/D卫星CZI影像,研究区域为黄海海域,重点关注山东南部、江苏东部近海,检测范围为31°~37°N,119°~125°E。本文使用CZI数据的L1C级产品,可在中国海洋卫星数据服务系统(https://osdds.nsoas.org.cn/#/)免费下载。通过ENVI软件无缝镶嵌工具对2~3景影像进行无缝拼接,重采样方法采用三次卷积法。接着进行辐射定标、快速大气校正等预处理,并利用预先标定的Shape文件,对陆地区域进行裁剪,仅留下海洋部分。
绿潮提取研究通常要求影像为薄云或无云状态,厚云影像中目标检测困难。因此,需要先去除厚云。对于HY-1C/D卫星CZI数据,由于仅有4个波段(band 1、band 2、band 3、band 4),可通过不同波段数值相减法构建6个新波段(b1 − b2、b1 − b3、b1 − b4、b2 − b3、b2 −b4、b3 − b4)。通过分析发现,厚云像元在band 3波段上数据分布单一,其数值大多在2 700~4 000左右,而绿潮像元在该波段上的数据分布为0~2 500。因此,可以设置band 3阈值为2 690,去除厚云像元(图1)。然而,该方法仅适用于亮度较高的厚云像元,对于薄云、厚云边缘和云斑等因素仍需后续处理。
在去除光谱特征较为单一的厚云像元后,海水影像中仍存在薄云、云斑耀斑等像元。为实现精细的绿潮区域提取,本文从无云、有云和耀斑影像中选择了绿潮像元与海水像元作为正负样本训练决策树,得到如表2所示的绿潮提取规则集。
为验证发现的绿潮提取规则的有效性,我们截取了无云、薄云、厚云等区域,采用该规则进行绿潮提取。若像元满足规则A,则分类为绿潮,否则分类为海水(颜色为黑色)。但此时的绿潮中仍有大量的云像元被误检测为绿潮,因此规则B–E为规则A的进一步细分。规则B为主要的绿潮提取规则,在无云情况下的提取结果如图2所示,绝大多数的绿潮像元均被正确分类,但在厚云边缘处仍存在被误检测为绿潮的像元;规则C的提取结果如图3红色部分所示,该规则将被厚云边缘覆盖的绿潮像元误检测为厚云边缘,需将其修正为绿潮;规则D的提取结果如图4蓝色部分所示,该规则将被薄云覆盖的绿潮像元误检测为薄云,需将其修正为绿潮;规则E的提取结果如图5橙色部分所示,结果中并不包含绿潮,而是厚云边缘处的薄云,但其与海水相接处存在零星的被误检测为绿潮的像元,需要将这些像元修正为云像元。因此,仅采用该规则并不能准确提取绿潮,仍存在一些误检测的情况,需制定策略修正错误类别。
根据实验结果,修正误检测像元可分为两种情况:(1)绿潮像元被误分类为云像元且临近绿潮像元。(2)云像元被误分为绿潮像元,通常分布离散且周围为厚云或海水。针对这些情况,引入滑动窗口概念,以3 × 3滑动窗口为例,若某非边缘像元为中心像元,则该像元和其周围8个像元构成滑动窗口,设计了以下5个步骤的修正策略。
策略1:绿潮一般以大型斑块存在于海面上,而不是点状分布。因此,如果中心像元为绿潮,其周围像元均为海水,则将该像元标记为待定类别,待后续修正策略判定是否为绿潮像元。
策略2:薄云覆盖下的绿潮斑块往往仍存在部分被正确分类的绿潮,依据二者的邻接关系提取出完整的绿潮斑块(图4)。因此,如果中心像元为薄云覆盖下的绿潮且滑窗内至少存在一个绿潮,则中心像元替换为绿潮。
策略3:在厚云边缘的薄云处,往往存在许多零星的被误检测为绿潮的像元(图5),同策略1相似,绿潮一般不以点状分布,且其与厚云邻接时往往呈现斑块状(图2)。因此,如果中心像元为绿潮且滑窗内至多两个绿潮:(1)如果滑窗内存在厚云或厚云边缘的薄云,中心像元标记为待定类别;(2)滑窗内云像元总数大于绿潮像元总数,即零星的绿潮像元被云像元围住的情况下,中心像元标记为待定类别。
策略4:经前3个修正策略作用,许多云覆盖下的绿潮得以修正,绿潮斑块更加完整,依据绿潮与云邻接时的往往呈现斑块状而非点状的分布规律,如果中心像元为薄云覆盖下的绿潮或厚云边缘覆盖下的绿潮且滑窗内至少存在一个绿潮,则中心像元替换为绿潮。
策略5:同策略4的原则相似,在绿潮斑块更加完整的情况下,如果中心像元为待定类别且滑窗内至少存在一个绿潮,则中心像元替换为绿潮。
错误类别修正过程即为采用滑动窗口的图像遍历过程,具体遍历方式如图6所示,首先按行正向遍历后再次逆向遍历,然后按列正向遍历后再次逆向遍历,每个策略反复执行,当本轮结果与上一轮结果完全相同时,进行下一策略。以图6b为例,假设图像尺寸为(n + 2) × (n + 2),数字代表滑窗内中心像元的序号,按行正向遍历的顺序即为(1, 2, 3, …, n, n + 1, …, 2n, 2n + 1, …, n2 n, n2 n + 1, …, n2),一次完整遍历(图6d)结束后,检测当前图像与遍历前图像的差异,若无差异则进入下一步骤;若仍存在差异,则重复本步骤。
使用本文提出的绿潮自动提取方法在复杂天气情况下的区域进行实验,得到的结果(图7c)表明:错误类别修正后的结果带有一定的预测性,如图7a1中下方绿潮斑块,可明显看出其被云覆盖的部分是绿潮像元,然而发现的绿潮提取规则无法将其分类为绿潮(图7b1蓝色部分),经修正,被误检测为薄云或被厚云边缘覆盖的绿潮像元得以纠正,同时保持了绿潮斑块的完整性;并且由于策略3的限制,在绿潮与厚云的接缝处(如图7b4红色与白色接缝处)这种纠正过程会及时停止,不会过度纠正;同时,无云条件下已正确提取的绿潮并不会受影响,修正前后绿潮斑块并无差异(图7b3图7c3)。因此本方法减轻了云像元对于绿潮像元提取的影响,使得分类结果更加接近有云天气下的绿潮真实分布情况。
本文提出的绿潮全自动提取方法的整个流程如图8所示,主要包括数据预处理、基于规则集的海水、绿潮分类,以及基于5种策略的后处理,用以修正错误类别,最终实现绿潮区域的高精度、全自动提取。其中,数据预处理包括图像镶嵌、辐射定标、大气校正、选取感兴趣区、去除厚云区域等步骤;使用余下的海水像元与绿潮像元训练决策树,得到初步区分海水、绿潮的规则集;为优化该规则,依据各规则所提取绿潮的特点,设计策略修正被误检测的绿潮像元,实现绿潮全自动提取。
本文以人工目视解译结果为标准,将本文方法与经典的归一化植被指数NDVI[11]、虚拟基线高度浮藻指数VB-FAH[15]等绿潮提取方法,以及ResNet50[19]、U-Net[21] 等深度学习方法进行对比实验,采用准确率(ACC)、Kappa系数、F1-Score、平均交并比(MIoU)作为精度评估的评价指标,验证本文提取方法的有效性。其中,评价指标分别定义如式(1)至式(6)。
准确率(ACC)表示预测正确的样本数量占全部样本的百分比,其定义为
$ \mathrm{ACC=\frac{TP+TN}{TP+TN+FP+FN}}, $
召回率(Recall)是指是在实际为正的样本中被预测为正样本的概率,其定义为
$ \mathrm{Recall=\frac{TP}{TP+FN}\text{,}} $
式中,TP表示一个样本的真实值为正例,预测值也为正例;FN表示一个样本的真实值为正例,但预测值为反例;FP表示一个样本的真实值为反例,但预测值为正例;TN表示一个样本的真实值为反例,预测值也为反例。本研究以人工目视解译结果作为真实值,算法提取结果作为预测值。
Kappa系数是基于混淆矩阵的计算、用于一致性检验的指标,可以用来衡量分类效果,其定义为
$ {\mathrm{Kappa}} = \frac{{{p_0} - {p_e}}}{{1 - {p_e}}}, $
式中,p0为对角线元素之和与整个矩阵元素之和的比值,即准确率(ACC);pe为所有类别分别对应的真实数量与预测数量的乘积之和与样本总数的平方的比值,其定义为
$ {p}_{e}=\frac{\displaystyle \sum _{i\;=\;1}第i行元素之和\;\times\; 第i列元素之和}{\left(\sum 矩阵所有元素\right)^{2}} . $
F1-Score为ACC与Recall的调和平均数,其定义为
$ {\mathrm{F}}1 {\text{-}} {\mathrm{Score}} = \frac{{2 \times {\mathrm{ACC}} \times {\mathrm{Recall}}}}{{{\mathrm{ACC}} + {\mathrm{Recall}}}} . $
平均交并比(MIoU)是各个类别预测结果与真实标签之间,交集与并集之间的比值之和的平均值,其定义为
$ {\mathrm{MI{o}U}}=\frac{1}{2}\left(\frac{{\mathrm{TP}}}{{\mathrm{TP+FP+FN}}}+\frac{{\mathrm{TN}}}{{\mathrm{TN+FN+FP}}}\right) . $
归一化植被指数NDVI的定义如下:
$ {\mathrm{NDVI }}= \frac{{{R_4} - {R_3}}}{{{R_4} + {R_3}}} , $
虚拟基线高度浮藻指数VB-FAH的定义如下:
$ {\text{VB-FAH}} = ({R_4} - {R_2}) + \frac{{({R_2} - {R_3}) \times ({\lambda _4} - {\lambda _2})}}{{2{\lambda _4} - {\lambda _3} - {\lambda _2}}} \text{,} $
式中,R4R3R2分别为近红外波段、红光波段、绿光波段的地表反射率,分别对应本文所用HY-1C/D卫星CZI数据的band 4、band 3、band 2;λ4λ3λ2分别为对应各波段的中心波长。
NDVI和VB-FAH方法提取绿潮的思路是增强绿潮区域的光谱值,并非自动提取方法,且阈值的选取往往因不同研究者所用的研究数据不同(即卫星、传感器、空间分辨率、图像预处理方式等不同)而有所差异,需要人为设定提取阈值。因此,如图9所示,本文选择一些绿潮与非绿潮区域的样本,NDVI方法设置绿潮提取阈值为0.24,VB-FAH方法设置绿潮提取阈值为211,高于设定阈值的区域为绿潮区域。
ResNet50与 U-Net网络均选用Adam作为优化器,交叉熵作为损失函数,单次传递给程序用以训练的参数个数(Batch Size)设置为1,训练轮次设置为40。ResNet50的初始学习率为0.005,U-Net的初始学习率为0.002。两个网络模型均在Tensorflow平台上实现,使用NVIDIA 3060Ti GPU进行实验。
实验数据选择2021年6月21日影像的区域1至区域15、2021年6月30日影像的区域16至区域20、2021年5月25日的区域21至区域25(图10)。为评估本文方法在不同条件下的提取效果,选取了厚云(指厚度较大、颜色单一的云层,通常为白色或灰色。厚云常常会遮挡太阳和天空,导致天气阴沉、昏暗。在卫星遥感图像中,厚云通常呈现出明显的白色或灰白色的云层,比较容易识别,见图11a)、薄云(指厚度较薄、颜色较浅的云层,通常为白色或灰色,常常呈现出细碎的条纹状或絮状。薄云通常会出现在低空或高空,对太阳光的遮挡程度较小,但仍然会影响天气的明亮度和舒适度,见图12a)、无云(指天空中没有云层遮挡的状态,通常是晴朗的天气,阳光充足,天空呈现出蔚蓝色或深蓝色。在卫星遥感图像中,无云通常呈现出蓝色或深蓝色的天空,见图13a)、云斑(指太阳光照射在云层上形成的阴影或云层之间的缝隙,在海面上呈现出碎片状的影子或颜色较深的区域,通常出现在云层较为厚重的区域,见图14a)、耀斑(太阳光入射至海面所形成的强烈的反射辐射或强风天气下阳光照射至海浪边缘所形成的闪光点,通常在阳光照射强烈的时候出现,见图15a)5种情况,如图10所示,共25景影像,提取效果如图11图15所示,精度评估结果如表3表4所示。
结果表明:(1)在厚云区域,NDVI方法和VB-FAH方法的绿潮提取效果并不理想,误检测现象严重;深度学习方法在没有云覆盖的情况下,对于绿潮的提取效果较好,但是对于存在云覆盖的绿潮像元,提取效果不佳;本文方法可以较好地纠正被云覆盖的绿潮像元。(2)在薄云区域,NDVI方法和VB-FAH方法因受到云像元的影响仍存在误检测;深度学习方法难以提取被云覆盖的绿潮因此存在绿潮斑块破碎现象,其提取结果并不充分;本文方法在修正错误类别前受薄云影响最为严重,经修正,该影响得以有效消除,提取的绿潮斑块更为完整。(3)在无云区域,5种方法的分类效果相似,但NDVI方法与本文方法均存在一定的“过度提取”现象(如图13中区域11、区域13)。笔者认为这种情况可能与绿潮斑块的中心部分与边缘部分的厚度与密度不同相关。较为明显的绿潮斑块中心部分较高,而亮度低的边缘部分可能仍漂浮着厚度较小、密度较低的绿潮并与海水混合在一起。故在遥感影像中,人工提取绿潮时难以直接确定其是否属于绿潮像元,而本文方法与对植被敏感的NDVI方法却可以将其检测到。(4)在云斑区域,NDVI方法和VB-FAH方法表现出差异性。NDVI方法存在将云像元误检测为绿潮的情况;而VB-FAH方法则表现为许多绿潮像元没有被提取到的“不充分提取”情况;深度学习方法中仍存在因难以提取云覆盖情况下的绿潮像元导致的绿潮斑块破碎的问题,相比之下本文方法更为优异。(5)在耀斑区域,NDVI方法提取结果基本包含真实绿潮像元,但有许多非绿潮像元也被判定为绿潮像元;其他方法多表现为“不充分提取”,其中本文方法“不充分提取”程度最小。笔者认为,耀斑区域中存在强烈的太阳耀光干扰,其经海水镜面反射产生的强辐射信号远高于海水或绿潮本身,因此导致影像与真实情况存在偏差,对人工标定绿潮和自动提取绿潮的准确性造成影响。从精度评估数据上来看,不同区域、不同指标上本文方法的精度均高于其他方法,ACC与F1-Score指标均达到0.97以上,Kappa系数与MIoU指标均达到0.81以上。可以认为,在绿潮自动提取工作中,本文方法相较于传统方法与深度学习方法,性能更为优越,尤其是在对云覆盖与耀斑的抗干扰性方面有较大提升。
本文以我国海洋水色HY-1C/D卫星多光谱影像为数据源,发现并给出了适用于各种复杂状况的HY-1C/D影像绿潮自动提取规则,以该规则为基础,探究绿潮提取中各种错误类别像元的分布特点,设计了5种高效的错误类别修正策略。该方法在中国近海的绿潮提取与监测任务中表现出以下优势:(1)相较于受云影响较大的NDVI、VB-FAH等传统方法,本文方法能够很好地减轻厚云、薄云、云斑等影响;(2)对于被云层部分覆盖的绿潮斑块,通过分析滑动窗口内中心像元与其邻接像元的分布情况来对中心像元的变动做出决策,带有一定的预测性,分类结果更加接近真实的绿潮分布情况;(3)在存在厚云、云斑、耀斑等大量影响因素的情况下,本文方法仍能表现出较高的精度,可以更加快速准确地完成绿潮提取任务。
此外,研究结果表明,绿潮斑块的形态差异使得基于光学遥感数据的绿潮提取任务具有一定的不确定性,较大斑块提取的不确定性较低,而零星的小型斑块不确定性较高。本研究认为这种不确定性可能来自绿潮斑块中心与边缘部分的厚度差异,这种差异在光学影像中尚无法精确呈现。在厚云覆盖下的多光谱影像中,厚云下的绿潮区域的信息基本完全丧失,提取出的结果与真实的绿潮分布情况存在较大误差,但考虑到绿潮漂移扩散路径与海洋流场具有较强的相关性,因此在今后的研究中,我们将结合绿潮斑块的形态学差异与时间序列下绿潮的分布信息,降低绿潮提取任务的不确定性,填补厚云覆盖下绿潮分布范围的空缺,从而获得更高精度的绿潮分布情况。
致谢: HY-1C/D卫星数据获取自网站:https://osdds.nsoas.org.cn。本文作者感谢国家卫星海洋应用中心提供的数据支持。
  • 国家自然科学基金项目(62172247);山东省重点研发计划重大科技创新工程项目(2019JZZY020101)。
参考文献 引证文献
排序方式:
1
Hiraoka M, Ohno M, Kawaguchi S, et al. Crossing test among floating Ulva thalli forming ‘green tide’ in Japan[J]. Hydrobiologia, 2004, 512(1): 239−245.
2
衣立, 张苏平, 殷玉齐. 2009年黄海绿潮浒苔爆发与漂移的水文气象环境[J]. 中国海洋大学学报, 2010, 40(10): 15−23.
Yi Li, Zhang Suping, Yin Yuqi. Influnce of environmental hydro-meteorological conditions to Enteromorpha prolifera blooms in Yellow Sea, 2009[J]. Periodical of Ocean University of China, 2010, 40(10): 15−23.
3
Liu Dongyan, Keesing J K, He Peimin, et al. The world’s largest macroalgal bloom in the Yellow Sea, China: formation and implications[J]. Estuarine, Coastal and Shelf Science, 2013, 129: 2−10.
4
Hu Lianbo, Zeng Kan, Hu Chuanmin, et al. On the remote estimation of Ulva prolifera areal coverage and biomass[J]. Remote Sensing of Environment, 2019, 223: 194−207.
5
Qi Lin, Hu Chuanmin, Xing Qianguo, et al. Long-term trend of Ulva prolifera blooms in the western Yellow Sea[J]. Harmful Algae, 2016, 58: 35−44.
6
Qi Lin, Hu Chuanmin. To what extent can Ulva and Sargassum be detected and separated in satellite imagery?[J]. Harmful Algae, 2021, 103: 102001.
7
Hu Lianbo, Hu Chuanmin, He Mingxia. Remote estimation of biomass of Ulva prolifera macroalgae in the Yellow Sea[J]. Remote Sensing of Environment, 2017, 192: 217−227.
8
Hu Chuanmin, Li Daqiu, Chen Changsheng, et al. On the recurrent Ulva prolifera blooms in the Yellow Sea and East China Sea[J]. Journal of Geophysical Research: Oceans, 2010, 115(C5): C05017.
9
Xing Qianguo, Hu Chuanmin, Tang Danling, et al. World’s largest macroalgal blooms altered phytoplankton biomass in summer in the Yellow Sea: satellite observations[J]. Remote Sensing, 2015, 7(9): 12297−12313.
10
Nelson T A, Haberlin K, Nelson A V, et al. Ecological and physiological controls of species composition in green macroalgal blooms[J]. Ecology, 2008, 89(5): 1287−1298.
11
Hu Chuanmin, He Mingxia. Origin and offshore extent of floating algae in Olympic sailing area[J]. Eos, Transactions American Geophysical Union, 2008, 89(33): 302−303.
12
施英妮, 石立坚, 夏明, 等. HJ-1A/1B星CCD传感器数据在黄东海浒苔监测中的应用[J]. 遥感信息, 2012, 27(2): 47−50.
Shi Yingni, Shi Lijian, Xia Ming, et al. The application of HJ-1A/1B’s CCD data to entermorpha prolifera monitoring over the Yellow Sea and East Sea[J]. Remote Sensing Information, 2012, 27(2): 47−50.
13
Hu Chuanmin. A novel ocean color index to detect floating algae in the global oceans[J]. Remote Sensing of Environment, 2009, 113(10): 2118−2129.
14
Son Y B, Min J E, Ryu J H. Detecting massive green algae (Ulva prolifera) blooms in the Yellow Sea and East China Sea using geostationary ocean color imager (GOCI) data[J]. Ocean Science Journal, 2012, 47(3): 359−375.
15
Xing Qianguo, Hu Chuanmin. Mapping macroalgal blooms in the Yellow Sea and East China Sea using HJ-1 and Landsat data: Application of a virtual baseline reflectance height technique[J]. Remote Sensing of Environment, 2016, 178: 113−126.
16
薛瑞, 吴孟泉, 刘杨, 等. 基于HJ-1A/1B的2014年黄海海域浒苔灾害时空分布[J]. 海洋科学, 2016, 40(7): 115−123.
Xue Rui, Wu Mengquan, Liu Yang, et al. Spatial and temporal variability of Ulva prolifera in the Yellow Sea, China in 2014[J]. Marine Sciences, 2016, 40(7): 115−123.
17
曾韬, 刘建强. “北京一号”小卫星在青岛近海浒苔灾害监测中的应用[J]. 遥感信息, 2009, 24(3): 34−37.
Zeng Tao, Liu Jianqiang. The application of Beijing-1 micro satellite data to algae disaster monitoring in the sea of Qingdao[J]. Remote Sensing Information, 2009, 24(3): 34−37.
18
刘锦超, 刘建强, 丁静, 等. HY-1C卫星CZI载荷的黄海绿潮提取研究[J]. 海洋学报, 2022, 44(5): 1−11.
Liu Jinchao, Liu Jianqiang, Ding Jing, et al. A refined imagery algorithm to extract green tide in the Yellow Sea from HY-1C satellite CZI measurements[J]. Haiyang Xuebao, 2022, 44(5): 1−11.
19
He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, USA: IEEE, 2016: 770−778.
20
Badrinarayanan V, Kendall A, Cipolla R. SegNet: a deep convolutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12): 2481−2495.
21
Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation[C]//18th International Conference on Medical Image Computing and Computer-Assisted Intervention. Munich: Springer, 2015: 234−241.
22
Cui Binge, Zhang Haoqing, Jing Wei, et al. SRSe-net: super-resolution-based semantic segmentation network for green tide extraction[J]. Remote Sensing, 2022, 14(3): 710.
23
Yu Haifei, Wang Changying, Li Jinhua, et al. Automatic extraction of green tide from GF-3 SAR images based on feature selection and deep learning[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021, 14: 10598−10613.
24
Wang Zhongyuan, Fang Zhixiang, Liang Jianfeng, et al. Estimating Ulva prolifera green tides of the Yellow Sea through ConvLSTM data fusion[J]. Environmental Pollution, 2023, 324: 121350.
25
Shang Weitao, Gao Zhiqiang, Gao Meng, et al. Monitoring green tide in the Yellow Sea using high-resolution imagery and deep learning[J]. Remote Sensing, 2023, 15(4): 1101.
26
Shi Wei, Wang Menghua. Green macroalgae blooms in the Yellow Sea during the spring and summer of 2008[J]. Journal of Geophysical Research: Oceans, 2009, 114(C12): C12010.
27
Wang Mengqiu, Hu Chuanmin. Mapping and quantifying Sargassum distribution and coverage in the Central West Atlantic using MODIS observations[J]. Remote Sensing of Environment, 2016, 183: 350−367.
28
苗珊珊. “长征”二号C运载火箭成功发射“海洋”一号C卫星[J]. 中国航天, 2018(9): 26.
Miao Shanshan. The CZ-2C carrier rocket successfully launched the HY-1C satellite[J]. Aerospace China, 2018(9): 26.
29
刘建强, 蒋兴伟, 王丽丽, 等. 海洋一号C、D卫星组网观测与应用[J]. 卫星应用, 2021(9): 19−26.
Liu Jianqiang, Jiang Xingwei, Wang Lili, et al. Monitoring and application of the HY-1C and D satellites constellation[J]. Satellite Application, 2021(9): 19−26.
30
李岩松, 赵慧洁, 李娜, 等. 基于中红外偏振的海面太阳耀光背景下的目标探测[J]. 中国激光, 2022, 49(19): 1910004.
Li Yansong, Zhao Huijie, Li Na, et al. Detection of marine targets covered in sun glint based on mid-infrared polarization[J]. Chinese Journal of Lasers, 2022, 49(19): 1910004.
31
Wang Chen, Zhang Huaguo, Xu Qing, et al. Inversion of the refractive index of marine spilled oil using multi-angle sun glitter images acquired by the ASTER sensor[J]. Remote Sensing of Environment, 2022, 275: 113019.
32
刘建强, 曾韬, 梁超, 等. 海洋一号C卫星在自然灾害监测中的应用[J]. 卫星应用, 2020(6): 26−34.
Liu Jianqiang, Zeng Tao, Liang Chao, et al. Application of HY-1C satellite in natural disaster monitoring[J]. Satellite Application, 2020(6): 26−34.
2023年第45卷第10期
PDF下载
149
57
引用本文
BibTeX
文章信息
doi: 10.12284/hyxb2023151
  • 接收时间:2023-02-15
  • 首发时间:2025-12-28
  • 出版时间:2023-10-01
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2023-02-15
  • 修回日期:2023-06-13
基金
国家自然科学基金项目(62172247);山东省重点研发计划重大科技创新工程项目(2019JZZY020101)。
作者信息
    1 青岛大学 计算机科学技术学院,山东 青岛, 266071

通讯作者:

*王常颖(1980—),副教授,主要从事海洋复杂性与数据挖掘研究。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/hyxb/CN/10.12284/hyxb2023151
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏