Article(id=1253997367842246810, tenantId=1146029695717560320, journalId=1251234736404742242, issueId=1253997366797857283, articleNumber=null, orderNo=null, doi=10.3969/j.issn.1672-2337.2025.05.003, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1730995200000, receivedDateStr=2024-11-08, revisedDate=1735315200000, revisedDateStr=2024-12-28, acceptedDate=null, acceptedDateStr=null, onlineDate=1776905870638, onlineDateStr=2026-04-23, pubDate=null, pubDateStr=null, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1776905870638, onlineIssueDateStr=2026-04-23, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1776905870638, creator=13041195026, updateTime=1776905870638, updator=13041195026, issue=Issue{id=1253997366797857283, tenantId=1146029695717560320, journalId=1251234736404742242, year='2025', volume='23', issue='5', pageStart='473', pageEnd='590', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1776905870387, creator=13041195026, updateTime=1777355497251, updator=13041195026, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1255883238690702048, tenantId=1146029695717560320, journalId=1251234736404742242, issueId=1253997366797857283, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1255883238690702049, tenantId=1146029695717560320, journalId=1251234736404742242, issueId=1253997366797857283, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=491, endPage=502, ext={EN=ArticleExt(id=1253997368093905052, articleId=1253997367842246810, tenantId=1146029695717560320, journalId=1251234736404742242, language=EN, title=Sea-Surface Small Target Detection Based on Joint Graph Features in Dual Channels, columnId=null, journalTitle=Radar Science and Technology, columnName=null, runingTitle=null, highlight=null, articleAbstract=

Currently, sea-surface small targets have become the focus and difficulty of marine radar detection. The existing detection methods are limited to the use of unilateral information such as radar echo amplitude or spectrum, making it difficult to effectively detect small targets. Thus, a sea-surface small target detection method using dual-channel joint graph features (DC-JGF) is proposed in this paper. Firstly, the time-domain phase sequence and the frequencydomain amplitude sequence are extracted from the radar complex echo sequence to generate the time-frequency domain dual channel. In each channel, graphs are generated separately to provide rich correlation information. Secondly, in the time domain channel, the largest and the second-largest eigenvalues of graph Laplacian matrix are fused as the first feature to evaluate the graph density. In the frequency domain channel, by extracting non-zero elements from the diagonal of the degree matrix, the entropy value is calculated as the second feature to measure the dispersion of the vertex distribution of the graph. Then, the two features are used as detection statistics to determine whether they fall within the decision region given by the convex hull algorithm with target guidance. The detection results are obtained. Finally, experimental results using measured data demonstrate that the proposed detector can achieve robust and efficient detection performance in complex detection environments.

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, 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=Sainan SHI, Sutong JIANG, Jiajun WANG, Tao LI), CN=ArticleExt(id=1253997379045232953, articleId=1253997367842246810, tenantId=1146029695717560320, journalId=1251234736404742242, language=CN, title=基于双通道图特征联合的海面小目标检测, columnId=0, journalTitle=雷达科学与技术, columnName=, runingTitle=null, highlight=null, articleAbstract=

目前,海面小目标已成为海洋雷达探测的重点和难点对象。现有的检测方法局限于雷达回波幅度或频谱的单方面信息的使用,难以有效探测小目标。为此,本文提出一种基于双通道图特征联合的海面小目标检测方法。首先,从雷达复回波序列中提取时域相位序列和频域幅度序列,生成时域-频域双通道。在每个通道,分别生成连通图,为检测提供丰富的相关性。其次,在时域通道,通过分析图拉普拉斯矩阵特征值的差异性,选取最大和次大特征值的融合值作为第一特征,衡量图的连通密度。在频域通道,通过提取度矩阵的对角线非零元素,计算其熵值作为第二特征,衡量图顶点分布的分散度。然后,将两个特征作为检测统计量,并判断是否落在具有目标引导的凸包算法决定的判决区域内,获得检测结果。最后,使用实测数据的实验结果表明,所提出的检测器在复杂探测环境下能够获得稳健且高效的检测性能。

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=Ydkf0DhX3vmuFTOI0VK6XQ==, magXml=ZUxcZgIj1Ko25mF5IefBeA==, pdfUrl=null, pdf=FJgbtpTzksdbqfmXxjEPNw==, pdfFileSize=6432631, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=2fKTJxFG6AuW8+rl01c9qA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=nOyq693TsYVSIzVHyslieg==, mapNumber=null, authorCompany=null, fund=null, authors=

施赛楠 女,博士,副教授,主要研究方向为雷达信号处理、微弱目标检测、气象雷达探测。

姜苏桐 女,硕士,主要研究方向为海面目标检测与识别。

汪佳俊 男,硕士,主要研究方向为无人机检测、智能网络检测。

李焘 女,博士,副教授,主要研究方向为SAR图像处理、目标检测与识别。

, authorsList=施赛楠, 姜苏桐, 汪佳俊, 李焘)}, authors=[Author(id=1253999614907061240, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, orderNo=0, 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=1253999615179691005, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, authorId=1253999614907061240, language=EN, stringName=Sainan SHI, firstName=Sainan, middleName=null, lastName=SHI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China
2.Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1253999615548788736, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, authorId=1253999614907061240, language=CN, stringName=施赛楠, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.南京信息工程大学电子与信息工程学院,江苏南京 210044
2.南京信息工程大学江苏省大气环境与装备技术协同创新中心,江苏南京 210044, bio={"content":"

施赛楠 女,博士,副教授,主要研究方向为雷达信号处理、微弱目标检测、气象雷达探测。

"}, bioImg=null, bioContent=

施赛楠 女,博士,副教授,主要研究方向为雷达信号处理、微弱目标检测、气象雷达探测。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1253999612554056676, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=1., ext=[AuthorCompanyExt(id=1253999612570833894, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612554056676, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China), AuthorCompanyExt(id=1253999612579222503, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612554056676, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.南京信息工程大学电子与信息工程学院,江苏南京 210044)]), AuthorCompany(id=1253999612663108584, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=2., ext=[AuthorCompanyExt(id=1253999612675691497, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612663108584, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China), AuthorCompanyExt(id=1253999612684080107, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612663108584, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.南京信息工程大学江苏省大气环境与装备技术协同创新中心,江苏南京 210044)])]), Author(id=1253999615699783684, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, orderNo=1, 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=1253999615834001416, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, authorId=1253999615699783684, language=EN, stringName=Sutong JIANG, firstName=Sutong, middleName=null, lastName=JIANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1253999615993384971, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, authorId=1253999615699783684, language=CN, stringName=姜苏桐, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.南京信息工程大学电子与信息工程学院,江苏南京 210044, bio={"content":"

姜苏桐 女,硕士,主要研究方向为海面目标检测与识别。

"}, bioImg=null, bioContent=

姜苏桐 女,硕士,主要研究方向为海面目标检测与识别。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1253999612554056676, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=1., ext=[AuthorCompanyExt(id=1253999612570833894, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612554056676, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China), AuthorCompanyExt(id=1253999612579222503, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612554056676, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.南京信息工程大学电子与信息工程学院,江苏南京 210044)])]), Author(id=1253999616085659662, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, 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=1253999616207294482, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, authorId=1253999616085659662, language=EN, stringName=Jiajun WANG, firstName=Jiajun, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1253999616333123603, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, authorId=1253999616085659662, language=CN, stringName=汪佳俊, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.南京信息工程大学电子与信息工程学院,江苏南京 210044, bio={"content":"

汪佳俊 男,硕士,主要研究方向为无人机检测、智能网络检测。

"}, bioImg=null, bioContent=

汪佳俊 男,硕士,主要研究方向为无人机检测、智能网络检测。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1253999612554056676, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=1., ext=[AuthorCompanyExt(id=1253999612570833894, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612554056676, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China), AuthorCompanyExt(id=1253999612579222503, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612554056676, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.南京信息工程大学电子与信息工程学院,江苏南京 210044)])]), Author(id=1253999616509284374, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, 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=1253999616685445146, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, authorId=1253999616509284374, language=EN, stringName=Tao LI, firstName=Tao, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, 4, address=3.School of Automation, Hangzhou Dianzi University, Hangzhou 310018, China
4.Nanjing Research Institute of Electronics Technology, Nanjing 210013, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1253999616760942620, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, authorId=1253999616509284374, language=CN, stringName=李焘, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, 4, address=3.杭州电子科技大学自动化学院,浙江 杭州 310018
4.南京电子技术研究所,江苏南京 210013, bio={"content":"

李焘 女,博士,副教授,主要研究方向为SAR图像处理、目标检测与识别。

"}, bioImg=null, bioContent=

李焘 女,博士,副教授,主要研究方向为SAR图像处理、目标检测与识别。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1253999612788937709, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=3., ext=[AuthorCompanyExt(id=1253999612793132014, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612788937709, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.School of Automation, Hangzhou Dianzi University, Hangzhou 310018, China), AuthorCompanyExt(id=1253999612801520623, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612788937709, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.杭州电子科技大学自动化学院,浙江 杭州 310018)]), AuthorCompany(id=1253999613028013042, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=4., ext=[AuthorCompanyExt(id=1253999613048984563, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999613028013042, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Nanjing Research Institute of Electronics Technology, Nanjing 210013, China), AuthorCompanyExt(id=1253999613053178868, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999613028013042, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.南京电子技术研究所,江苏南京 210013)])])], keywords=[Keyword(id=1253999616983240735, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, orderNo=1, keyword=sea clutter), Keyword(id=1253999617050349601, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, orderNo=2, keyword=target detection), Keyword(id=1253999617163595811, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, orderNo=3, keyword=dual channels), Keyword(id=1253999617276842020, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, orderNo=4, keyword=connected graph), Keyword(id=1253999617402671141, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, orderNo=5, keyword=feature fusion), Keyword(id=1253999617536888871, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, orderNo=1, keyword=海杂波), Keyword(id=1253999619160084520, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, orderNo=2, keyword=目标检测), Keyword(id=1253999619273330731, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, orderNo=3, keyword=双通道), Keyword(id=1253999619411742765, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, orderNo=4, keyword=连通图), Keyword(id=1253999619579514927, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, orderNo=5, keyword=特征联合)], refs=[Reference(id=1253999625027915879, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2024, volume=22, issue=1, pageStart=57, pageEnd=62, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=王威, 杨勇, 韩静雯, journalName=雷达科学与技术, refType=null, unstructuredReference=王威,杨勇,韩静雯.基于极化联合特征值的雷达弱小目标检测方法[J].雷达科学与技术,2024,22(1):57-62., articleTitle=基于极化联合特征值的雷达弱小目标检测方法, refAbstract=null), Reference(id=1253999625145356393, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2024, volume=13, issue=3, pageStart=501, pageEnd=524, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=陈小龙, 何肖阳, 邓振华, journalName=雷达学报, refType=null, unstructuredReference=陈小龙,何肖阳,邓振华,.雷达微弱目标智能化处理技术与应用[J].雷达学报,2024,13(3):501-524., articleTitle=雷达微弱目标智能化处理技术与应用, refAbstract=null), Reference(id=1253999625237631083, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2013, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=2, authorNames=WARD K D, TOUGH R J A, WATTS S, journalName=Sea Clutter: Scattering, the K Distribution and Radar Performance, refType=null, unstructuredReference=WARD K D, TOUGH R J A, WATTS S. Sea Clutter: Scattering, the K Distribution and Radar Performance[M].UK: Institute of Engineering Technology, 2013., articleTitle=null, refAbstract=null), Reference(id=1253999625329905772, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2024, volume=22, issue=1, pageStart=77, pageEnd=86, url=null, language=null, rfNumber=[4], rfOrder=3, authorNames=施赛楠, 高季娟, 李东宸, journalName=雷达科学与技术, refType=null, unstructuredReference=施赛楠,高季娟,李东宸.基于多维矩特征联合的K分布杂波参数估计[J].雷达科学与技术,2024,22(1):77-86., articleTitle=基于多维矩特征联合的K分布杂波参数估计, refAbstract=null), Reference(id=1253999625401208941, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2006, volume=54, issue=1, pageStart=136, pageEnd=143, url=null, language=null, rfNumber=[5], rfOrder=4, authorNames=HU Jing, TUNG Wenwen, GAO Jianbo, journalName=IEEE Trans on Antennas and Propagation, refType=null, unstructuredReference=HU Jing, TUNG Wenwen, GAO Jianbo. Detection of Low Observable Targets Within Sea Clutter by Structure Function Based Multifractal Analysis[J]. IEEE Trans on Antennas and Propagation, 2006, 54(1):136-143., articleTitle=Detection of Low Observable Targets Within Sea Clutter by Structure Function Based Multifractal Analysis, refAbstract=null), Reference(id=1253999625464123502, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2020, volume=101, issue=null, pageStart=102707, pageEnd=null, url=null, language=null, rfNumber=[6], rfOrder=5, authorNames=GUO Zixun, SHUI Penglang, BAI Xiaohui, journalName=Digital Signal Processing, refType=null, unstructuredReference=GUO Zixun, SHUI Penglang, BAI Xiaohui. Small Target Detection in Sea Clutter Using All-dimensional Hurst Exponents of Complex Time Sequence[J]. Digital Signal Processing, 2020, 101:102707., articleTitle=Small Target Detection in Sea Clutter Using All-dimensional Hurst Exponents of Complex Time Sequence, refAbstract=null), Reference(id=1253999625615118447, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2014, volume=50, issue=2, pageStart=1416, pageEnd=1430, url=null, language=null, rfNumber=[7], rfOrder=6, authorNames=SHUI Penglang, LI Dongchen, XU Shuwen, journalName=IEEE Trans on Aerospace & Electronic Systems, refType=null, unstructuredReference=SHUI Penglang, LI Dongchen, XU Shuwen. Tri-Feature-Based Detection of Floating Small Targets in Sea Clutter[J]. IEEE Trans on Aerospace & Electronic Systems, 2014, 50(2):1416-1430., articleTitle=Tri-Feature-Based Detection of Floating Small Targets in Sea Clutter, refAbstract=null), Reference(id=1253999625686421616, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2022, volume=44, issue=3, pageStart=63, pageEnd=69, url=null, language=null, rfNumber=[8], rfOrder=7, authorNames=施赛楠, 杨静, 董泽远, journalName=现代雷达, refType=null, unstructuredReference=施赛楠,杨静,董泽远.基于高维特征域随机森林的海面小目标检测[J].现代雷达,2022,44(3):63-69., articleTitle=基于高维特征域随机森林的海面小目标检测, refAbstract=null), Reference(id=1253999625753530481, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2023, volume=16, issue=null, pageStart=9422, pageEnd=9436, url=null, language=null, rfNumber=[9], rfOrder=8, authorNames=BAI Xiaohui, XU Shuwen, ZHU Jianan, journalName=IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, refType=null, unstructuredReference=BAI Xiaohui, XU Shuwen, ZHU Jianan, et al. Floating Small Target Detection in Sea Clutter Based on Multifeature Angle Variance[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2023, 16:9422-9436., articleTitle=Floating Small Target Detection in Sea Clutter Based on Multifeature Angle Variance, refAbstract=null), Reference(id=1253999625829027954, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2019, volume=16, issue=8, pageStart=1225, pageEnd=1229, url=null, language=null, rfNumber=[10], rfOrder=9, authorNames=LI Yuzhou, XIE Pengcheng, TANG Zechen, journalName=IEEE Geoscience and Remote Sensing Letters, refType=null, unstructuredReference=LI Yuzhou, XIE Pengcheng, TANG Zechen, et al. SVM-Based Sea-Surface Small Target Detection: A False-Alarm-Rate-Controllable Approach[J]. IEEE Geoscience and Remote Sensing Letters, 2019, 16(8): 1225-1229., articleTitle=SVM-Based Sea-Surface Small Target Detection: A False-Alarm-Rate-Controllable Approach, refAbstract=null), Reference(id=1253999625980022899, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2019, volume=26, issue=6, pageStart=793, pageEnd=797, url=null, language=null, rfNumber=[11], rfOrder=10, authorNames=ZHOU Hongkuan, JIANG Tao, journalName=IEEE Geoscience and Remote Sensing Letters, refType=null, unstructuredReference=ZHOU Hongkuan, JIANG Tao. Decision Tree Based Sea-Surface Weak Target Detection with False Alarm Rate Controllable[J]. IEEE Geoscience and Remote Sensing Letters, 2019, 26(6):793-797., articleTitle=Decision Tree Based Sea-Surface Weak Target Detection with False Alarm Rate Controllable, refAbstract=null), Reference(id=1253999626084880500, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2020, volume=56, issue=6, pageStart=4947, pageEnd=4964, url=null, language=null, rfNumber=[12], rfOrder=11, authorNames=GUO Zixun, SHUI Penglang, journalName=IEEE Trans on Aerospace and Electronic Systems, refType=null, unstructuredReference=GUO Zixun, SHUI Penglang. Anomaly-Based Sea-Surface Small Target Detection Using K-Nearest-Neighbour Classification[J]. IEEE Trans on Aerospace and Electronic Systems, 2020, 56(6):4947-4964., articleTitle=Anomaly-Based Sea-Surface Small Target Detection Using K-Nearest-Neighbour Classification, refAbstract=null), Reference(id=1253999626202321013, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2017, volume=21, issue=1, pageStart=20, pageEnd=23, url=null, language=null, rfNumber=[13], rfOrder=12, authorNames=YAN Kun, WU Hsiao-Chun, XIAO Hailin, journalName=IEEE Communications Letters, refType=null, unstructuredReference=YAN Kun,WU Hsiao-Chun, XIAO Hailin, et al. Novel Robust Band-Limited Signal Detection Approach Using Graphs[J]. IEEE Communications Letters, 2017, 21(1):20-23., articleTitle=Novel Robust Band-Limited Signal Detection Approach Using Graphs, refAbstract=null), Reference(id=1253999626349121654, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2021, volume=43, issue=11, pageStart=3185, pageEnd=3192, url=null, language=null, rfNumber=[14], rfOrder=13, authorNames=时艳玲, 姚婷婷, 郭亚星, journalName=电子与信息学报, refType=null, unstructuredReference=时艳玲,姚婷婷,郭亚星.基于图连通密度的海面漂浮小目标检测[J].电子与信息学报,2021,43(11):3185-3192., articleTitle=基于图连通密度的海面漂浮小目标检测, refAbstract=null), Reference(id=1253999626449784951, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2023, volume=45, issue=5, pageStart=1567, pageEnd=1574, url=null, language=null, rfNumber=[15], rfOrder=14, authorNames=许述文, 焦银萍, 白晓惠, journalName=电子与信息学报, refType=null, unstructuredReference=许述文,焦银萍,白晓惠,.基于频域多通道图特征感知的海面小目标检测[J].电子与信息学报,2023, 45(5):1567-1574., articleTitle=基于频域多通道图特征感知的海面小目标检测, refAbstract=null), Reference(id=1253999626512699512, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2024, volume=46, issue=7, pageStart=2712, pageEnd=2720, url=null, language=null, rfNumber=[16], rfOrder=15, authorNames=许述文, 何绮, 茹宏涛, journalName=电子与信息学报, refType=null, unstructuredReference=许述文,何绮,茹宏涛.基于无监督图互信息最大化的海面小目标异常检测[J].电子与信息学报,2024,46 (7):2712-2720., articleTitle=基于无监督图互信息最大化的海面小目标异常检测, refAbstract=null), Reference(id=1253999628093952121, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2024, volume=12, issue=1, pageStart=86, pageEnd=119, url=null, language=null, rfNumber=[17], rfOrder=16, authorNames=MACHARETE R R, DEL-VECCHIO R R, TEIXEIRA H, journalName=Special Matrices, refType=null, unstructuredReference=MACHARETE R R, DEL-VECCHIO R R, TEIXEIRA H, et al. A Laplacian Eigenbasis for Threshold Graphs[J]. Special Matrices, 2024, 12(1):86-119., articleTitle=A Laplacian Eigenbasis for Threshold Graphs, refAbstract=null), Reference(id=1253999628194615418, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2023, volume=15, issue=4, pageStart=429, pageEnd=438, url=null, language=null, rfNumber=[18], rfOrder=17, authorNames=施赛楠, 姜丽, 曹鼎, journalName=南京信息工程大学学报(自然科学版), refType=null, unstructuredReference=施赛楠,姜丽,曹鼎,.基于频域相对样本熵的海面小目标特征检测[J].南京信息工程大学学报(自然科学版),2023,15(4):429-438., articleTitle=基于频域相对样本熵的海面小目标特征检测, refAbstract=null), Reference(id=1253999628261724283, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2024-08-15, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=18, authorNames=Cognitive Systems Laboratory, journalName=null, refType=null, unstructuredReference=Cognitive Systems Laboratory. IPIX Radar Database[EB/OL].[2024-08-15]. http://soma.mcmaster.ca//ipix.php., articleTitle=IPIX Radar Database, refAbstract=null), Reference(id=1253999628333027452, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, doi=null, pmid=null, pmcid=null, year=2023, volume=12, issue=2, pageStart=456, pageEnd=469, url=null, language=null, rfNumber=[20], rfOrder=19, authorNames=关键, 刘宁波, 王国庆, journalName=雷达学报, refType=null, unstructuredReference=关键,刘宁波,王国庆,.雷达对海探测试验与目标特性数据获取——海上目标双极化多海况散射特性数据集[J].雷达学报,2023,12(2):456-469., articleTitle=雷达对海探测试验与目标特性数据获取——海上目标双极化多海况散射特性数据集, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1253999612554056676, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=1., ext=[AuthorCompanyExt(id=1253999612570833894, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612554056676, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China), AuthorCompanyExt(id=1253999612579222503, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612554056676, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.南京信息工程大学电子与信息工程学院,江苏南京 210044)]), AuthorCompany(id=1253999612663108584, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=2., ext=[AuthorCompanyExt(id=1253999612675691497, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612663108584, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China), AuthorCompanyExt(id=1253999612684080107, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612663108584, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.南京信息工程大学江苏省大气环境与装备技术协同创新中心,江苏南京 210044)]), AuthorCompany(id=1253999612788937709, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=3., ext=[AuthorCompanyExt(id=1253999612793132014, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612788937709, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.School of Automation, Hangzhou Dianzi University, Hangzhou 310018, China), AuthorCompanyExt(id=1253999612801520623, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999612788937709, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.杭州电子科技大学自动化学院,浙江 杭州 310018)]), AuthorCompany(id=1253999613028013042, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, xref=4., ext=[AuthorCompanyExt(id=1253999613048984563, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999613028013042, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.Nanjing Research Institute of Electronics Technology, Nanjing 210013, China), AuthorCompanyExt(id=1253999613053178868, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, companyId=1253999613028013042, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4.南京电子技术研究所,江苏南京 210013)])], figs=[ArticleFig(id=1253999619885699121, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=Qhv0Bkz+v0Iq2qceOwXtIg==, figureFileBig=2fKTJxFG6AuW8+rl01c9qA==, tableContent=null), ArticleFig(id=1253999619961196595, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图1, caption=双通道幅度和相位的波动性分析, figureFileSmall=Qhv0Bkz+v0Iq2qceOwXtIg==, figureFileBig=2fKTJxFG6AuW8+rl01c9qA==, tableContent=null), ArticleFig(id=1253999620112191541, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=WmGs+Mh4AuAJ5yk6tatj1g==, figureFileBig=+Ss/f5CEPltZC79690da9g==, tableContent=null), ArticleFig(id=1253999620217049143, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图2, caption=DC-JGF检测器流程图, figureFileSmall=WmGs+Mh4AuAJ5yk6tatj1g==, figureFileBig=+Ss/f5CEPltZC79690da9g==, tableContent=null), ArticleFig(id=1253999620330295353, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=XUKhOWbuPm8RdDRDG7cbNw==, figureFileBig=kUjRc7mQjlY/ROn8wTKhIQ==, tableContent=null), ArticleFig(id=1253999620481290299, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图3, caption=海杂波和含目标序列的连通图生成演示, figureFileSmall=XUKhOWbuPm8RdDRDG7cbNw==, figureFileBig=kUjRc7mQjlY/ROn8wTKhIQ==, tableContent=null), ArticleFig(id=1253999620632285245, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=OU9dXbkCWg+NCRwV63qdDg==, figureFileBig=QfrWLsSWsda8CKg515uixQ==, tableContent=null), ArticleFig(id=1253999620720365631, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图4, caption=拉普拉斯矩阵特征值分析, figureFileSmall=OU9dXbkCWg+NCRwV63qdDg==, figureFileBig=QfrWLsSWsda8CKg515uixQ==, tableContent=null), ArticleFig(id=1253999620846194753, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=HE447fYkJfcyVElFohUUPA==, figureFileBig=nsVXLOzgNRRxiSqsXiBvSg==, tableContent=null), ArticleFig(id=1253999621005578307, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图5, caption=时域幅度和相位序列的FF分布, figureFileSmall=HE447fYkJfcyVElFohUUPA==, figureFileBig=nsVXLOzgNRRxiSqsXiBvSg==, tableContent=null), ArticleFig(id=1253999621135601733, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=HdmWPySdAJfKESfvhrIN6A==, figureFileBig=foHi9dcAjZ+Bj2XSVDO69Q==, tableContent=null), ArticleFig(id=1253999621261430854, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图6, caption=频域幅度和相位序列的EF分布, figureFileSmall=HdmWPySdAJfKESfvhrIN6A==, figureFileBig=foHi9dcAjZ+Bj2XSVDO69Q==, tableContent=null), ArticleFig(id=1253999621395648584, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=CtKCLv4E769DH+2BWrfnKA==, figureFileBig=MqONOmcngmLlsjFpkTIlAw==, tableContent=null), ArticleFig(id=1253999621492117578, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图7, caption=具有目标引导的二维凸包判决区域演示, figureFileSmall=CtKCLv4E769DH+2BWrfnKA==, figureFileBig=MqONOmcngmLlsjFpkTIlAw==, tableContent=null), ArticleFig(id=1253999621630529612, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=sJ0jSvRRdRvCfhpud794DA==, figureFileBig=B8vK+9QECaaXxoP0ad1k3Q==, tableContent=null), ArticleFig(id=1253999621789913165, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图8, caption=10组数据在4种极化下的平均信杂比, figureFileSmall=sJ0jSvRRdRvCfhpud794DA==, figureFileBig=B8vK+9QECaaXxoP0ad1k3Q==, tableContent=null), ArticleFig(id=1253999621898965072, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=XY9Bek6MTkoZqv5hngNwmg==, figureFileBig=6RwWrq3L/cpNEEbWc89Vxg==, tableContent=null), ArticleFig(id=1253999621978656850, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图9, caption=量化点数对性能的影响, figureFileSmall=XY9Bek6MTkoZqv5hngNwmg==, figureFileBig=6RwWrq3L/cpNEEbWc89Vxg==, tableContent=null), ArticleFig(id=1253999623580880980, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=aQuY4EHPHRZEjPjfDbUT2w==, figureFileBig=UP5F88ZER+lumocUuPZXxQ==, tableContent=null), ArticleFig(id=1253999623723487318, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图10, caption=不同数据下5种检测器性能对比, figureFileSmall=aQuY4EHPHRZEjPjfDbUT2w==, figureFileBig=UP5F88ZER+lumocUuPZXxQ==, tableContent=null), ArticleFig(id=1253999623870287959, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=vtewbBcgvk7GmD8urBNCAQ==, figureFileBig=evFblGuIJCdWkGCue7AEkQ==, tableContent=null), ArticleFig(id=1253999624000311384, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图11, caption=实测数据检测结果, figureFileSmall=vtewbBcgvk7GmD8urBNCAQ==, figureFileBig=evFblGuIJCdWkGCue7AEkQ==, tableContent=null), ArticleFig(id=1253999624105168986, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=OP2HoYD68wpunsMjrAlBZQ==, figureFileBig=9tLTcYS6McqhOpe2RP4cGw==, tableContent=null), ArticleFig(id=1253999624205832284, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=图12, caption=实测数据幅度分布和测试目标, figureFileSmall=OP2HoYD68wpunsMjrAlBZQ==, figureFileBig=9tLTcYS6McqhOpe2RP4cGw==, tableContent=null), ArticleFig(id=1253999624302301278, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
检测器N=256(0.25s)N=512(0.512s)N=1024(1.024s)
Hurst指数检测器[5]0.169/0.254/0.244/0.1470.248/0.351/0.351/0.2900.324/0.469/0.469/0.396
GS检测器[13]0.169/0.254/0.244/0.1470.190/0.367/0.399/0.3290.173/0.376/0.373/0.336
CDG检测器[14]0.207/0.300/0.297/0.0860.291/0.297/0.396/0.1360.327/0.472/0.466/0.200
三特征检测器[7]0.507/0.605/0.595/0.5110.534/0.622/0.619/0.5180.588/0.674/0.667/0.542
FF检测器0.425/0.449/0.470/0.4340.558/0.604/0.599/0.5060.628/0.670/0.667/0.562
EF检测器0.260/0.412/0.411/0.2580.266/0.388/0.411/0.2730.254/0.470/0.462/0.293
DC-JGF检测器0.519/0.608/0.588/0.5100.584/0.656/0.663/0.5800.663/0.739/0.732/0.650
), ArticleFig(id=1253999624390381664, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=表1, caption=

不同脉冲累积数下4种极化的平均检测概率

, figureFileSmall=null, figureFileBig=null, tableContent=
检测器N=256(0.25s)N=512(0.512s)N=1024(1.024s)
Hurst指数检测器[5]0.169/0.254/0.244/0.1470.248/0.351/0.351/0.2900.324/0.469/0.469/0.396
GS检测器[13]0.169/0.254/0.244/0.1470.190/0.367/0.399/0.3290.173/0.376/0.373/0.336
CDG检测器[14]0.207/0.300/0.297/0.0860.291/0.297/0.396/0.1360.327/0.472/0.466/0.200
三特征检测器[7]0.507/0.605/0.595/0.5110.534/0.622/0.619/0.5180.588/0.674/0.667/0.542
FF检测器0.425/0.449/0.470/0.4340.558/0.604/0.599/0.5060.628/0.670/0.667/0.562
EF检测器0.260/0.412/0.411/0.2580.266/0.388/0.411/0.2730.254/0.470/0.462/0.293
DC-JGF检测器0.519/0.608/0.588/0.5100.584/0.656/0.663/0.5800.663/0.739/0.732/0.650
), ArticleFig(id=1253999624545570915, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
检测器N=256
(0.128s)
N=512
(0.256s)
N=1024
(0.512s)
Hurst指数检测器[5]0.0330.0740.109
GS检测器[13]0.0980.1840.312
CDG检测器[14]0.0370.1020.141
三特征检测器[7]0.2710.3790.422
DC-JGF检测器0.5550.5980.688
), ArticleFig(id=1253999624730120292, tenantId=1146029695717560320, journalId=1251234736404742242, articleId=1253997367842246810, language=CN, label=表2, caption=

实测数据下5种检测器的检测概率

, figureFileSmall=null, figureFileBig=null, tableContent=
检测器N=256
(0.128s)
N=512
(0.256s)
N=1024
(0.512s)
Hurst指数检测器[5]0.0330.0740.109
GS检测器[13]0.0980.1840.312
CDG检测器[14]0.0370.1020.141
三特征检测器[7]0.2710.3790.422
DC-JGF检测器0.5550.5980.688
)], attaches=null, journal=Journal(id=1251231495537340518, delFlag=0, nameCn=雷达科学与技术, nameEn=Radar Science and Technology, nameHistory1=null, nameHistory2=null, issn=1672-2337, eissn=, cn=34-1264/TN, coden=null, periodic=1, language=CN, oaType=1, 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=, officePhone=, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=OjNlWJvYThs/yBUoSYSlfQ==, journalPrice=null, startedYear=null, abbrevIsoEn=Radar Science and Technology, journalRemark=null, publicationField=null, createdTime=1776246435296, updatedTime=1776397629352, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=R, firstLetterEn=R, subjectCode=Engineering, subjectName=工程, subjectCodeEn=Engineering, subjectNameEn=null, picCn=OjNlWJvYThs/yBUoSYSlfQ==, picEn=4rjTZHs3SxCXTtsO7C1btA==, jcr=null, cjcr=null, exts=[JournalExt(id=1251865649480414057, language=CN, name=雷达科学与技术, 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=, createdTime=1776397629375, updatedTime=1776397629375, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://radarst.ijournal.cn/ldkxyjs/ch/author/login.aspx, submissionEditorUrl=http://radarst.ijournal.cn/ldkxyjs/ch/login.aspx, submissionReviewUrl=http://radarst.ijournal.cn/ldkxyjs/ch/auditor/login.aspx, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1251865649518162794, language=EN, name=Radar Science and Technology, 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=, createdTime=1776397629384, updatedTime=1776397629384, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://radarst.ijournal.cn/ldkxyjs/ch/author/login.aspx, submissionEditorUrl=http://radarst.ijournal.cn/ldkxyjs/ch/login.aspx, submissionReviewUrl=http://radarst.ijournal.cn/ldkxyjs/ch/auditor/login.aspx, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1251234736404742242, websiteList=[Website(id=1251257283527786566, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1251234736404742242, 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/ldkxyjs/CN, language=CN, createTime=1776252583629, createBy=18614031015, updateTime=1776253947851, updateBy=18614031015, name=雷达科学与技术-中文, tplId=1146099689490845704, title=雷达科学与技术, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1251263103057474233, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=articleTextType, value=kx, createTime=1776253971113, updateTime=1776253971113, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103036502710, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=banner, value=null, createTime=1776253971108, updateTime=1776253971108, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103078445756, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=grayFlag, value=0, createTime=1776253971118, updateTime=1776253971118, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103023919797, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=logo, value=https://castjournals.cast.org.cn/joweb/ldkxyjs/CN/file/pic?fileId=4UasbXLM5gAH92UTSWwtsQ==, createTime=1776253971105, updateTime=1776253971105, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103091028670, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=minRunFlag, value=0, createTime=1776253971121, updateTime=1776253971121, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103053279928, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/ldkxyjs/CN/file/pic, createTime=1776253971112, updateTime=1776253971112, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103086834365, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=silenceFlag, value=0, createTime=1776253971120, updateTime=1776253971120, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103044891319, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1776253971110, updateTime=1776253971110, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103065862842, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=themeColor, value=null, createTime=1776253971115, updateTime=1776253971115, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263103070057147, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283527786566, code=themeStyle, value=null, createTime=1776253971116, updateTime=1776253971116, creator=18614031015, updator=18614031015)]), Website(id=1251257283615866956, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1251234736404742242, 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/ldkxyjs/EN, language=EN, createTime=1776252583650, createBy=18614031015, updateTime=1776253944503, updateBy=18614031015, name=雷达科学与技术-英文, tplId=1146101810881728533, title=Radar Science and Technology, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1251263075903553911, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=articleTextType, value=kx, createTime=1776253964639, updateTime=1776253964639, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075886776692, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=banner, value=null, createTime=1776253964635, updateTime=1776253964635, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075924525434, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=grayFlag, value=0, createTime=1776253964644, updateTime=1776253964644, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075878388083, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=logo, value=https://castjournals.cast.org.cn/joweb/ldkxyjs/EN/file/pic?fileId=4UasbXLM5gAH92UTSWwtsQ==, createTime=1776253964633, updateTime=1776253964633, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075937108348, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=minRunFlag, value=0, createTime=1776253964647, updateTime=1776253964647, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075899359606, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/ldkxyjs/EN/file/pic, createTime=1776253964638, updateTime=1776253964638, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075928719739, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=silenceFlag, value=0, createTime=1776253964645, updateTime=1776253964645, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075890970997, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1776253964636, updateTime=1776253964636, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075907748216, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=themeColor, value=null, createTime=1776253964640, updateTime=1776253964640, creator=18614031015, updator=18614031015), WebsiteProps(id=1251263075916136825, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283615866956, code=themeStyle, value=null, createTime=1776253964642, updateTime=1776253964642, creator=18614031015, updator=18614031015)])], journalTitle=雷达科学与技术, weixinUrl=null, journalUrl=http://radarst.ijournal.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Radar Science and Technology, journalPhotoCn=OjNlWJvYThs/yBUoSYSlfQ==, journalPhotoEn=4rjTZHs3SxCXTtsO7C1btA==, journalFirstLetter=R, 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/ldkxyjs/CN/10.3969/j.issn.1672-2337.2025.05.003, detailUrlEn=https://castjournals.cast.org.cn/joweb/ldkxyjs/EN/10.3969/j.issn.1672-2337.2025.05.003, pdfUrlCn=https://castjournals.cast.org.cn/joweb/ldkxyjs/CN/PDF/10.3969/j.issn.1672-2337.2025.05.003, pdfUrlEn=https://castjournals.cast.org.cn/joweb/ldkxyjs/EN/PDF/10.3969/j.issn.1672-2337.2025.05.003, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于双通道图特征联合的海面小目标检测
收藏切换
PDF下载
施赛楠 1, 2 , 姜苏桐 1 , 汪佳俊 1 , 李焘 3, 4
雷达科学与技术 | 2025,23(5): 491-502
收起
收藏切换
雷达科学与技术 | 2025, 23(5): 491-502
基于双通道图特征联合的海面小目标检测
全屏
施赛楠1, 2, 姜苏桐1, 汪佳俊1, 李焘3, 4
作者信息
  • 1.南京信息工程大学电子与信息工程学院,江苏南京 210044
  • 2.南京信息工程大学江苏省大气环境与装备技术协同创新中心,江苏南京 210044
  • 3.杭州电子科技大学自动化学院,浙江 杭州 310018
  • 4.南京电子技术研究所,江苏南京 210013
  • 施赛楠 女,博士,副教授,主要研究方向为雷达信号处理、微弱目标检测、气象雷达探测。

    姜苏桐 女,硕士,主要研究方向为海面目标检测与识别。

    汪佳俊 男,硕士,主要研究方向为无人机检测、智能网络检测。

    李焘 女,博士,副教授,主要研究方向为SAR图像处理、目标检测与识别。

Sea-Surface Small Target Detection Based on Joint Graph Features in Dual Channels
Sainan SHI1, 2, Sutong JIANG1, Jiajun WANG1, Tao LI3, 4
Affiliations
  • 1.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China
  • 2.Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China
  • 3.School of Automation, Hangzhou Dianzi University, Hangzhou 310018, China
  • 4.Nanjing Research Institute of Electronics Technology, Nanjing 210013, China
doi: 10.3969/j.issn.1672-2337.2025.05.003
文章导航
收藏切换

目前,海面小目标已成为海洋雷达探测的重点和难点对象。现有的检测方法局限于雷达回波幅度或频谱的单方面信息的使用,难以有效探测小目标。为此,本文提出一种基于双通道图特征联合的海面小目标检测方法。首先,从雷达复回波序列中提取时域相位序列和频域幅度序列,生成时域-频域双通道。在每个通道,分别生成连通图,为检测提供丰富的相关性。其次,在时域通道,通过分析图拉普拉斯矩阵特征值的差异性,选取最大和次大特征值的融合值作为第一特征,衡量图的连通密度。在频域通道,通过提取度矩阵的对角线非零元素,计算其熵值作为第二特征,衡量图顶点分布的分散度。然后,将两个特征作为检测统计量,并判断是否落在具有目标引导的凸包算法决定的判决区域内,获得检测结果。最后,使用实测数据的实验结果表明,所提出的检测器在复杂探测环境下能够获得稳健且高效的检测性能。

海杂波  /  目标检测  /  双通道  /  连通图  /  特征联合

Currently, sea-surface small targets have become the focus and difficulty of marine radar detection. The existing detection methods are limited to the use of unilateral information such as radar echo amplitude or spectrum, making it difficult to effectively detect small targets. Thus, a sea-surface small target detection method using dual-channel joint graph features (DC-JGF) is proposed in this paper. Firstly, the time-domain phase sequence and the frequencydomain amplitude sequence are extracted from the radar complex echo sequence to generate the time-frequency domain dual channel. In each channel, graphs are generated separately to provide rich correlation information. Secondly, in the time domain channel, the largest and the second-largest eigenvalues of graph Laplacian matrix are fused as the first feature to evaluate the graph density. In the frequency domain channel, by extracting non-zero elements from the diagonal of the degree matrix, the entropy value is calculated as the second feature to measure the dispersion of the vertex distribution of the graph. Then, the two features are used as detection statistics to determine whether they fall within the decision region given by the convex hull algorithm with target guidance. The detection results are obtained. Finally, experimental results using measured data demonstrate that the proposed detector can achieve robust and efficient detection performance in complex detection environments.

sea clutter  /  target detection  /  dual channels  /  connected graph  /  feature fusion
施赛楠, 姜苏桐, 汪佳俊, 李焘. 基于双通道图特征联合的海面小目标检测. 雷达科学与技术, 2025 , 23 (5) : 491 -502 . DOI: 10.3969/j.issn.1672-2337.2025.05.003
Sainan SHI, Sutong JIANG, Jiajun WANG, Tao LI. Sea-Surface Small Target Detection Based on Joint Graph Features in Dual Channels[J]. Radar Science and Technology, 2025 , 23 (5) : 491 -502 . DOI: 10.3969/j.issn.1672-2337.2025.05.003
海面小目标已成为海洋雷达探测的重点和难点对象,对海上航行安全、国防安全等具有重要的意义[1-2]。感兴趣的海面小目标主要有救生筏、海面漂浮物、小船等。由于小目标具有雷达截面积小、运动速度慢、机动性弱等典型特点,目标回波往往淹没在强杂波背景中。传统检测器单纯依靠回波能量或频谱往往性能损失严重,出现大量虚警点。因此,挖掘多域数据进行联合,是提升小目标检测的有效途径。
在雷达信号处理中,通常采用长时观测技术来提高信杂比。当观测时间长达几百毫秒至几秒时,动态海杂波呈现出非平稳、非均匀、非高斯性的复杂特性[3-4],小目标呈现出时变的多普勒偏移。此时,动态海面的非线性特性可采用分形理论描述。Hu等人[5]提出基于Hurst指数的海面目标检测方法,反映了动态海杂波幅度的分形特性。Guo等人[6]提出全域Hurst指数融合的检测方法,充分利用了复数回波、幅度、相位的分形特性。但是,基于分形的检测器一般需要几秒以上才能获得较好的性能。为此,Shui等人[7]从时域和频域提取3个特征,在三维特征空间中进行联合检测,开启了多维特征检测框架。随后,施赛楠等人[8]进一步联合时频域特征,提出基于随机森林的七维特征检测。Bai等人[9]又增加相位域的3个特征,提出基于角度异常检测的十一维特征检测。随着具有互补性特征的数目增加,检测器的性能也随之提升,同时带来了高维特征分类器设计的困难。一些学者引入了K近邻、一类支持向量机、决策树[10-12]等,试图通过调节参数控制虚警率,但结构复杂且计算量大。因此,综合考虑检测性能和计算复杂性,建议在低维设计更加有效的特征。
事实上,海杂波和目标回波序列之间存在较好的相关性,然而传统检测器中缺失了这种相关性。为此,Yan等人[13]首次将图方法引入信号检测中,通过构建频谱序列的连通图,充分获取数据间的相关性。并且,提出基于图谱(Graph on Spectrum,GS)的检测器,在连通图上提取拉普拉斯矩阵的次大特征值作为检测统计量,在高信杂比下具有较好的检测性能。随后,时艳玲等人[14]进一步考虑了连通图中各条边的连通次数,将拉普拉斯矩阵的最大特征值作为检测统计量,实现基于图连通密度(Connected Density of Graph,CDG)的小目标检测。在此基础上,许述文等人[15]融合了频域相对多普勒峰高特征[7]和图的连通密度[14],将两者的比值作为检测统计量。当目标落在主杂波带外且具有高信杂比时,该检测器具有较好的性能。此外,许述文等人[16]挖掘了8个特征之间的相关性信息,并采用图注意力网络实现检测。综上,可以发现海杂波和含目标回波主要差异性来源于时域幅值和频谱幅值,暂时缺乏对相位序列的应用。
为此,本文提出基于双通道图特征联合的海面小目标检测方法,充分挖掘雷达复时间序列的信息,以提高小目标检测性能。主要创新点体现在以下两个层面。在数据层面,采用时域-频域双通道,在保留频谱幅度信息的基础上,进一步增加时域相位信息。在检测层面,提取两个精心设计的特征,衡量图连通密度和图节点分散度。并结合具有目标引导的凸包算法,实现二维特征平面内的异常检测。
在雷达目标检测中,雷达发射机发送长度为N的连续脉冲信号。经过海面与目标的反射后,雷达通过接收天线接收回波。在接收机经过下变频、匹配滤波、脉冲压缩等处理后,进行回波中是否有目标信号的判决。此时,可以将判断雷达回波中是否存在目标的问题描述为以下的二元假设检验[5-9]
式中,zn)为待检测单元(Cell Under Test,CUT),cn)表示海杂波回波,sn)表示目标回波。如果CUT中仅有海杂波,判决为H0假设,即目标不存在;如果CUT含有目标回波,则判决为H1假设,即目标存在。
目前,基于图的目标检测方法仅使用频域幅度信息[13-15]。为了更多的挖掘到雷达回波的信息,本文采用时域-频域双通道数据,提取幅度和相位信息。在时域通道,计算复时间序列zn)的幅度序列和相位序列
式中,|·|表示取幅度,arg[·]表示取相位,Re[·]表示取实部,Im[·]表示取虚部。在频域通道,复回波时间序列zn)的多普勒频谱为
然后,按照式(2)计算频谱的幅度序列|Z(k)|和相位序列arg[Z(k)]。
下面,讨论如何选取通道的序列,保证后续获得更优的检测性能。考虑到图通过序列的前后关系生成,所以序列的波动性与图的性质密切相关。海杂波和含目标回波两者之间的波动性差异越大,反映在图的某些特性上差异也越大,越有利于检测。因此,引入变差系数(Coefficient of Variation,CV)衡量序列的波动程度,定义为标准差和均值的比值。图1画出了不同距离单元下的CV曲线,目标位于第8个距离单元,第7、9、10为影响单元,其余为海杂波单元。在图1(a)图1(b)中,时域幅度的CV值随机分布在14个距离单元内,目标单元和海杂波单元的CV值差异较小。然而,时域相位的CV值呈现规律性,目标单元和海杂波的CV值有显著的区分性。在图1(c)图1(d)中,海杂波单元和目标单元在频域幅度和频域相位上的变化规律相同,但两者在幅度上差异性更大。在频谱幅度上,目标单元和海杂波单元CV值差异性与信杂比密切相关,在高信杂比下两者的差异性更大。因此,采用CV值相差大的序列所生成的图具有更大的区分性。综上,选择时域相位序列与频域幅度序列,用于后续图生成。
针对雷达回波复序列,本文挖掘时域相位序列和频域幅度序列的特征,提出一种基于双通道图特征联合(Dual-Channel Joint Graph Features,DCJGF)的检测器。图2给出了DC-JGF检测器的流程图,包含检测分支和训练分支。在检测分支中,首先,CUT中N个连续脉冲回波转换为双通道序列,即时域的相位序列和频域的幅度序列。其次,对时域-频域双通道序列进行特征提取,所有的特征来源于各个通道的连通图。在时域通道,提取拉普拉斯矩阵的最大特征值和次大特征值的均值,即融合特征(Fused Feature,FF)。在频域通道,提取度矩阵的熵特征(Entropy Feature,EF)。然后,将两个特征构成一个二维的特征向量ξ,作为检测统计量。最后,在二维特征空间中,若ξ落在判决区域Ω内,则CUT为海杂波;反之,CUT中含有目标。在训练分支中,从大量的海杂波时间序列中获得M个特征向量,组成训练集合。然后,在给定的虚警率Pfa条件下,获取基于目标引导的快速凸包学习算法的判决区域,为检测分支提供门限信息。
首先,将序列xn)映射到(0,1)区间,得到归一化序列
式中,max表示取最大值,min表示取最小值。然后,对归一化序列U进行均匀量化,得到量化序列
式中,γ为量化点数。最后,对量化序列Q(n)生成连通图
式中:V={1,2,…,γ}为顶点集;E为边集,表示序列之间的连接关系。假设Qn)对应的量化值为2,Qn+1)对应的量化值为3,则增加一条由顶点2指向顶点3的有向边,当前权重为1。依次类推,可以得到γ个顶点之间的邻接矩阵
式中,ωij为权重,表示从第i顶点到第j个顶点连接的总次数。
图3演示了双通道中海杂波和含目标序列的图结构。设置序列长度N=512,量化数γ=10。图中顶点数为10个,带箭头的连接线表示图的有向边,数值代表连接次数。在时域通道,含目标图的边分布较为稀疏,图连通度较小。海杂波图的边分布较为稠密,接近全连接图,图连通度较大。在频域通道,含目标回波图的顶点分布极为不均衡,少数节点度高度集中,而大部分顶点度数极小。然而,海杂波图顶点分布相对更均衡,不存在孤立节点。这些特性反映了海杂波和含目标回波序列的差异性,在后续检测中具有重要的意义。
在连通图中,海杂波和含目标序列在连通密度和顶点分布上存在明显差异性。为了描述这种差异性,从邻接矩阵A出发,得到拉普拉斯矩阵[17]
式中,D为度矩阵,表明不同量化值顶点间的连接情况。度矩阵为对角矩阵,即
对角元素描述了对应顶点的度数。那么,拉普拉斯矩阵的对角元素Lii等于顶点i的度减去自环数,非对角元素Lijij)表示顶点ij的连接关系。这种构造方式不仅反映顶点与其邻居顶点间的关系,也反映了顶点和整体图的联系,适用于分析复杂网络结构。因此,拉普拉斯矩阵在图论、谱聚类、机器学习等方面有广泛的应用。
在时域通道,如图3(a)图3(b)所示,含目标相位序列的图连通密度小,而海杂波相位序列的图连通密度大。这种连通密度可以通过拉普拉斯矩阵的特征值描述。文献[13]将次大特征值作为检测统计量,而文献[14]将最大特征值作为检测统计量。事实上,拉普拉斯矩阵所有特征值都蕴含关于图连通密度的信息。为此,需要研究如何选取合适的特征值以及选取多少个特征值。
根据时域相位序列arg[xn)]生成连通图G1,计算拉普拉斯矩阵L。然后,计算拉普拉斯矩阵L的降序特征值
图4(a)中,海杂波的L矩阵特征值明显大于含目标回波的。这表明海杂波相位序列构成的图具有较大的连通密度,并且最大特征值最能反映自身序列的这一特性。然而,检测更加关注的是海杂波和含目标回波两者之间的差异性,差异性越大,越利于提高检测性能。因此,求解两者特征值的差值。可以发现,随着特征值逐渐下降,含目标回波和海杂波之间的差异性逐渐减小。并且,在第3个特征值后,两者的差异性显著减小。
接下来,有必要研究选取多少个特征值进行检测。在图4(b)中,横轴是融合特征值的个数,纵轴是均匀融合后的特征值。当融合个数为1,此时取最大特征值。由于特征值已按照从大到小排序,所以随着融合个数增加,融合的平均特征值逐步减少。这表明越大的特征值越重要,可以去除一部分小的特征值。在蓝色曲线上,当取融合个数为2时,海杂波和含目标回波两者的连通密度差异性最大。因此,将最大和次大特征值的平均值作为最终的融合特征(FF)
FF值越大,表明图的连通密度越大,图越稠密。
图5讨论了时域幅度和相位的图连通密度的差异性。在图5(a)中,对于时域幅度序列生成的图,含目标序列与海杂波序列的FF值有一定区分度,但在许多样本中两者值相近,甚至出现前者超过后者的,即特征混淆现象。这在检测中会出现高虚警率,导致检测概率下降。在图5(b)中,对于时域相位序列生成的图,含目标序列的FF值与杂波序列的FF值具有显著的区分性,仅在个别样本上出现特征混淆。这意味着时域相位更有利于提取FF,更能反映海杂波和含目标回波在连通密度上的差异性。
在频域通道,如图3(c)图3(d)所示,含目标幅度序列的顶点分布较为集中,而海杂波幅度序列的顶点分布较为分散。这种顶点分布分散度可以通过度矩阵的熵描述。在度矩阵中,对角线元素反映了顶点与其他顶点(包括自己)的连接次数。熵能够描述图顶点连接的混乱程度,在海面小目标检测中也有一定的应用[18]
假设度矩阵D对角线元素构成的向量为
那么,去除d中的零元素得到非零向量
γ0为矩阵dnz的长度,如果每个顶点都有连接到其他顶点的边,则γ0的值等于量化等级γ。然后,计算度矩阵的熵特征(EF)
EF值越大,表示顶点分布越混乱,连接分散且复杂,每个顶点的度(即与其他节点的连接数)有较大的变化。反之,EF值越小,表示顶点分布较为有序,顶点的连接较为集中。
图6分析了频域幅度序列和相位序列的EF分布情况。从整体上看,海杂波序列的EF值大于含目标序列的EF值,这表明海杂波频谱序列具有更大的离散度,而含目标频谱序列具有一定的规则性。在图6(a)中,从频域幅度中提取的含目标EF值主要分布在0~0.5之间,而海杂波的EF值波动在1上下。含目标序列和海杂波序列的特征混杂现象较少,两者界限清晰。在图6(b)中,从频域相位中提取的EF值,海杂波序列整体值在3.2附近,与含目标序列的EF值存在较多的重叠,难以区分海杂波和含目标回波。因此,在频域通道,选择幅度进行提取特征具有一定的合理性。
事实上,时域相位和频域幅度反映了雷达回波的不同域信息,可以进行联合检测,从而提高检测性能。因此,将双通道提取的FF和EF构成一个二维的特征向量
那么,式(1)中的检测问题,转换为二维特征空间的异常检测问题
式中,Ω表示判决区域,由虚警率Pfa决定。海杂波为正常样本,含目标回波为异常样本。当雷达开机并获得大量的海杂波序列样本后,判决区域可通过二维快速学习凸包算法获得。
在二维特征空间中,传统的快速学习凸包算法是按照面积最小为原则进行不断剔除虚警点。但是,这种方法没有考虑到含目标回波所在的位置,只保证去除距离中心最远的离散点,但不一定是目标的点。因此,本文将含目标回波的先验信息,作为引导信息,提出具有目标引导的凸包学习算法。具体算法步骤如下:
步骤1:输入训练集Δ和虚警率Pfa
按照图2中的训练分支,从大量海杂波序列中获取M个特征向量,构成训练集Δ={ξii=1,2,…,M}。根据给定的虚警率Pfa,确定虚警点数目为[M×Pfa],其中[·]是取整。
步骤2:设置目标引导点
从特征提取过程中,得到含目标回波的先验信息,即含目标回波的两个特征值都小于海杂波。因而,含目标回波的特征向量更容易聚集的区域为
选择目标可能区域的任意一点作为引导点,记为Oxy)。
步骤3:获取判决区域
首先,生成一个二维凸包,将训练集Δ全部包住,记顶点为q1q2,…,qQ。在Q个顶点中,寻找到目标引导点O的欧式距离最近的点
并将第j个顶点qj作为虚警点剔除,更新训练集Δ。然后,重复上述操作,继续删除虚警点,直到剔除所需求的虚警点,停止算法。
图7演示了二维特征平面上的判决区域获取过程,横轴为FF值,纵轴为EF值。海杂波训练集为10210个特征向量,虚警率Pfa=0.001。红色虚线框为目标引导区域,图中选取了原点作为引导点。黑色直线表示凸包判决区域,落在凸包外共10个虚警点。可以发现,所有的虚警点都集中在离目标引导区域最近的位置,也就是最可能判决为含目标的位置。虽然凸包的面积不是最小的,但是去除的虚警点是最准确的,更加符合实际雷达检测。
在实际探测环境中,雷达一旦开机,可获得大量当前杂波环境下的训练数据。凸包训练属于离线训练,不影响检测器的在线检测时间。当杂波环境发生改变时,需要重新启动凸包训练,更新判决区域,保证检测器对当前环境的实时感知。
在海面小目标检测中,1993年加拿大McMaster大学公开的IPIX雷达数据集[19]已成为公认的数据集。IPIX雷达具有HH、HV、VH、VV四种极化方式,工作在驻留模式,载波频率为9.39GHz,距离分辨率为30m,脉冲重复频率为1kHz,波束宽度0.9°。测试小目标为锚固定的包裹着金属丝网的塑料球,球直径1m,随海浪上下起伏。该数据的具体风速、风向、海况等环境参数可以参考文献[19]。每组数据共有14个距离单元,每个距离单元包含217个脉冲,即驻留时间为131s,海况包含2~4级。图8列出了10组数据在4种极化方式下的平均信杂比(Signal-to-Clutter Ratio,SCR)。交叉极化(HV、VH)的平均SCR要大于同相极化(HH、VV),VV极化下的平均SCR最小。从整体上来看,不同数据的平均SCR差异较大,取值范围从-2dB到18dB,具有较大的动态范围。同时,数据涉及了2级海况、3级海况和4级海况,可用于测试不同海况下的检测性能。
在连通图构建时,采用有限离散值对观测序列进行量化,势必会丢失部分信息。为了保证信息的完整性,量化点数γ越多越好。但随着γ增加,计算复杂度也相应增加,同时也会带来冗余信息。因此,有必要讨论合理选择量化点数。
设置序列长度N=512,结果如图9所示。可以发现,量化点数与检测概率的曲线有先上升后下降的过程,4种极化下呈现相同的现象。在HH极化下,当γ>10后,检测概率出现明显的下降趋势,随后保持平稳,这意味着太多的量化点数反而会带来冗余信息。对于交叉极化,γ在7~10之间具有较好的稳健性能。对于VV极化,γ在7~12之间性能最佳。综合考虑计算量、检测性能和极化方式,本文建议γ=10。在实际探测环境中,可根据多种因素进行折中选取。
下面,在不同实测数据下验证检测器的性能。设置脉冲累积数N=512,虚警率Pfa=10-3。为了保证样本数量,序列之间存在重叠,每段间隔为128。图10给出了5种检测器在不同数据和极化方式下的检测性能对比。在5种检测器中,基于Hurst指数检测器[5]、基于GS检测器[13]和基于CDG检测器[14]性能大幅度差于其余两种检测器。这是因为这3个检测器局限于单个特征,性能势必弱于多个特征联合。当观测时间为0.512s时,基于Hurst指数检测器[5]在HH、HV、VH、VV四种极化下的平均检测概率分别为0.248,0.351,0.351,0.290。Hurst指数可以反映时域幅度序列的分形特性,但需要观测时间达到几秒以上,其在短时间上,性能损失严重。基于GS检测器[13]提取了频域幅度拉普拉斯矩阵的次大特征值,平均检测概率分别为0.190,0.367,0.399,0.329。除了HH极化,其余极化方式下具有微弱的性能提升。相对于GS检测器[13],除了VV极化,基于CDG检测器[14]有一定提升,平均检测概率为0.291,0.297,0.396,0.136。这主要是考虑了连通图的顶点之间的连接次数,但这种提升性能有限,需要进一步挖掘不同通道内的信息。相对于这3种单特征检测器,基于三特征检测器[7]有了显著的性能提升,平均检测概率分别为0.534,0.622,0.619,0.518。这是因为联合了时域幅度1个特征和频域幅度2个特征,主要反映了海杂波和含目标回波在幅度和能量分布上的差异性。该检测器在平均信杂比高的数据下,具有很好的检测性能。从整体上来看,提出的DC-JGF检测器具有最佳的检测性能,检测概率分别为0.584,0.656,0.663,0.580。特别是,在平均SCR较小的情况下,比如数据#280和数据#310,该检测器仍获得最优检测器。相对于基于三特征检测器[7],提出的检测器在仅使用2个特征时仍有8.4%的性能提升。这是因为相位序列比幅度序列受SCR影响较小,在低SCR下仍能探测。所以,提出的DC-JGF检测器虽然维度上性能会略差于基于三特征检测器[7],但是两个精心设计的特征弥补了维度上的性能损失。
接下来,讨论脉冲累积数N对于检测性能的影响,如表1所示。设置N=256,512,1024(即观测时间为0.256,0.512,1.024s),虚警率Pfa=10-3,其他参数与前面实验保持一致。4个数值依次代表HH、HV、VH、VV极化下的10组数据平均检测概率。从表中可以归纳以下3个结论。第一,随着脉冲累积数增加,所有检测器的性能均有显著提升。第二,在所有检测器中,提出DC-JGF检测器的整体性能最佳。与基于三特征检测器[7]相比,在低观测时长时,两者检测性能接近。但是,随着脉冲累积数的上升,在N=1024时达到12.7%的性能提升。第三,分析所使用两个通道特征的性能。FF检测器和EF检测器只使用本文提出的单个特征进行检测。可以发现,时域相位特征整体性能优于频域幅度特征,且在所有观测时长上的性能均优于基于Hurst指数检测器[5]、基于GS检测器[13]、基于CDG检测器[14]。但是,在#26、#30、#31的VV极化下,EF检测器反而优于FF检测器,这表明这两个特征在复杂环境下可以进行互补。同时,这也证实了所提出特征的有效性,融合了较多的含目标和海杂波的差异性特性。由此可见,在复杂多变的实际探测环境中,本文提出的DC-JGF检测器可以获得高效且稳健的性能,特别是在观测时长达秒级时。
下面,使用1998年IPIX数据[19]验证检测器的性能。数据名称为19980204 163113 ANTSTEP,雷达载频为9.39GHz,距离分辨率为30m,脉冲重复频率为1kHz,工作在驻留模式下,波束宽度为0.9°。图11(a)给出了HV极化下的时间-距离幅度分布,一共28个距离单元,观测时间为60s。可以看到,海杂波幅度在时间-距离二维平面上呈现空时变特性,呈现出较强的非高斯特性。测试目标为漂浮的小船,位于第24个距离单元内。可以计算平均SCR为16.9dB,但目标单元具有较强的幅度起伏特性,这会增加检测的难度。图11(b)~(f)给出了5种检测器的检测结果,N=512,Pfa=10-3。当平均SCR较高时,基于Hurst指数检测器[5]、基于GS检测器[13]、基于CDG检测器[14]的检测概率分别为0.626,0.761,0.860,检测性能良好。此时,基于三特征检测器[7]的检测概率为0.880,没有太明显的性能提升。这说明在高SCR下,维度的增益不会太明显,也就是说更重要的是设计有效的特征。相对于基于三特征检测器[7],提出的DC-JGF检测器的检测概率为0.989,性能提升12.4%。这得益于相位序列和幅度序列双通道信息联合以及精心设计的两个特征。
最后,使用海军航空大学公开的雷达对海探测试验数据进行验证[20]。实验采用天奥SPPR50PHH雷达,工作在驻留模式,射频频率为9.3GHz,脉冲重复频率为2000Hz,距离分辨率为6m,采样点间隔2.5m。数据名20221112150043 stare HH,采用T2LFM信号,脉宽为8μs,水平波束宽度1.2°。图12(a)为脉冲-采样点幅度分布图,500个采样点,每个采样点包含217个脉冲。可以清晰看到,海杂波在时间上呈现空时纹理特性,且具有较强的非高斯特性。在图12(b)中,测试目标为钢材质浮标,浮体直径为2.4m,海面以上高度约4.1m,呈现锚定漂浮状态。测试目标位于498~506个采样点,平均SCR为4.5dB,海况等级为4级。考虑到目标占据多个采样点,将功率最大的第503个采样点作为目标所在位置,用于测试检测器的性能。
表2给出了5种检测器在不同累积脉冲数N下的检测概率,虚警率为Pfa=10-3。门限由蒙特卡洛试验10万次以上获得,保证达到给定虚警率。从表2中可知,基于Hurst指数检测器[5]的检测概率低于0.1,处于无法工作状态,这是因为观测时间为ms级别,并且SCR较低。基于GS检测器[13]和基于CDG检测器[14]的检测概率低于0.3,这是因为只采用了频域幅度数据,意味着检测器性能高度依赖于SCR。基于三特征检测器[7]虽然有3个特征的联合使用,但3个特征来自于时域幅度和频域幅度,同样性能与SCR密切相关。在低SCR条件下,本文提出的DC-JGF检测器取得最佳的检测效果。这主要得益于时域通道的相位序列的使用,与SCR相关性较小。
本文主要研究了一种基于双通道图特征联合的海面小目标检测方法。在现有图目标检测仅利用频谱幅度的基础上,增加了时域相位序列信息,并精心设计了两个特征,反映了海杂波和含目标回波的图连通密度和图顶点的差异性。同时,发展了具有目标引导的快速凸包学习算法,有助于获得更加准确的判决区域。最后,IPIX雷达实测数据验证了所提算法的高效性和稳健性,可用于实际复杂多变的杂波环境。
参考文献 引证文献
排序方式:
[1]
王威,杨勇,韩静雯.基于极化联合特征值的雷达弱小目标检测方法[J].雷达科学与技术,2024,22(1):57-62.
[2]
陈小龙,何肖阳,邓振华,.雷达微弱目标智能化处理技术与应用[J].雷达学报,2024,13(3):501-524.
[3]
WARD K D, TOUGH R J A, WATTS S. Sea Clutter: Scattering, the K Distribution and Radar Performance[M].UK: Institute of Engineering Technology, 2013.
[4]
施赛楠,高季娟,李东宸.基于多维矩特征联合的K分布杂波参数估计[J].雷达科学与技术,2024,22(1):77-86.
[5]
HU Jing, TUNG Wenwen, GAO Jianbo. Detection of Low Observable Targets Within Sea Clutter by Structure Function Based Multifractal Analysis[J]. IEEE Trans on Antennas and Propagation, 2006, 54(1):136-143.
[6]
GUO Zixun, SHUI Penglang, BAI Xiaohui. Small Target Detection in Sea Clutter Using All-dimensional Hurst Exponents of Complex Time Sequence[J]. Digital Signal Processing, 2020, 101:102707.
[7]
SHUI Penglang, LI Dongchen, XU Shuwen. Tri-Feature-Based Detection of Floating Small Targets in Sea Clutter[J]. IEEE Trans on Aerospace & Electronic Systems, 2014, 50(2):1416-1430.
[8]
施赛楠,杨静,董泽远.基于高维特征域随机森林的海面小目标检测[J].现代雷达,2022,44(3):63-69.
[9]
BAI Xiaohui, XU Shuwen, ZHU Jianan, et al. Floating Small Target Detection in Sea Clutter Based on Multifeature Angle Variance[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2023, 16:9422-9436.
[10]
LI Yuzhou, XIE Pengcheng, TANG Zechen, et al. SVM-Based Sea-Surface Small Target Detection: A False-Alarm-Rate-Controllable Approach[J]. IEEE Geoscience and Remote Sensing Letters, 2019, 16(8): 1225-1229.
[11]
ZHOU Hongkuan, JIANG Tao. Decision Tree Based Sea-Surface Weak Target Detection with False Alarm Rate Controllable[J]. IEEE Geoscience and Remote Sensing Letters, 2019, 26(6):793-797.
[12]
GUO Zixun, SHUI Penglang. Anomaly-Based Sea-Surface Small Target Detection Using K-Nearest-Neighbour Classification[J]. IEEE Trans on Aerospace and Electronic Systems, 2020, 56(6):4947-4964.
[13]
YAN Kun,WU Hsiao-Chun, XIAO Hailin, et al. Novel Robust Band-Limited Signal Detection Approach Using Graphs[J]. IEEE Communications Letters, 2017, 21(1):20-23.
[14]
时艳玲,姚婷婷,郭亚星.基于图连通密度的海面漂浮小目标检测[J].电子与信息学报,2021,43(11):3185-3192.
[15]
许述文,焦银萍,白晓惠,.基于频域多通道图特征感知的海面小目标检测[J].电子与信息学报,2023, 45(5):1567-1574.
[16]
许述文,何绮,茹宏涛.基于无监督图互信息最大化的海面小目标异常检测[J].电子与信息学报,2024,46 (7):2712-2720.
[17]
MACHARETE R R, DEL-VECCHIO R R, TEIXEIRA H, et al. A Laplacian Eigenbasis for Threshold Graphs[J]. Special Matrices, 2024, 12(1):86-119.
[18]
施赛楠,姜丽,曹鼎,.基于频域相对样本熵的海面小目标特征检测[J].南京信息工程大学学报(自然科学版),2023,15(4):429-438.
[19]
Cognitive Systems Laboratory. IPIX Radar Database[EB/OL].[2024-08-15]. http://soma.mcmaster.ca//ipix.php.
[20]
关键,刘宁波,王国庆,.雷达对海探测试验与目标特性数据获取——海上目标双极化多海况散射特性数据集[J].雷达学报,2023,12(2):456-469.
2025年第23卷第5期
PDF下载
55
25
引用本文
BibTeX
文章信息
doi: 10.3969/j.issn.1672-2337.2025.05.003
  • 接收时间:2024-11-08
  • 首发时间:2026-04-23
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-11-08
  • 修回日期:2024-12-28
基金
作者信息
    1.南京信息工程大学电子与信息工程学院,江苏南京 210044
    2.南京信息工程大学江苏省大气环境与装备技术协同创新中心,江苏南京 210044
    3.杭州电子科技大学自动化学院,浙江 杭州 310018
    4.南京电子技术研究所,江苏南京 210013
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/ldkxyjs/CN/10.3969/j.issn.1672-2337.2025.05.003
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