Article(id=1153809125017375052, tenantId=1146029695717560320, journalId=1152916057816748034, issueId=1153809123054440769, articleNumber=null, orderNo=null, doi=10.3969/j.issn.2095–1469.2024.05.07, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1690732800000, receivedDateStr=2023-07-31, revisedDate=1694707200000, revisedDateStr=2023-09-15, acceptedDate=null, acceptedDateStr=null, onlineDate=1753019132140, onlineDateStr=2025-07-20, pubDate=null, pubDateStr=null, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1753019132140, onlineIssueDateStr=2025-07-20, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1753019132140, creator=13701087609, updateTime=1753019132140, updator=13701087609, issue=Issue{id=1153809123054440769, tenantId=1146029695717560320, journalId=1152916057816748034, year='2024', volume='14', issue='5', pageStart='745', pageEnd='919', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=0, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1753019131673, creator=13701087609, updateTime=1757481656846, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1172526310380417245, tenantId=1146029695717560320, journalId=1152916057816748034, issueId=1153809123054440769, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1172526310384611550, tenantId=1146029695717560320, journalId=1152916057816748034, issueId=1153809123054440769, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=812, endPage=820, ext={EN=ArticleExt(id=1153809125382279501, articleId=1153809125017375052, tenantId=1146029695717560320, journalId=1152916057816748034, language=EN, title=Lane Line Detection Based on Attention Mechanism and Line Anchor Information Transmission, columnId=1153809124719579465, journalTitle=Chinese Journal of Automotive Engineering, columnName=Intelligent & Connected Technologies Section/Editor-in-Chief: GAO Zhenhai, runingTitle=null, highlight=null, articleAbstract=

Lane line detection is a key technology in the field of autonomous driving, and it currently faces many challenges. The sparsity of the lane line supervision signal, as well as factors such as occlusion and shadows in complex scenes, can affect detection accuracy and realtime performance. Based on this, this paper proposes a lane line detection model that integrates the CBAM attention mechanism and a line anchor feature aggregation module. The proposed algorithm achieves an accuracy of 96.19% and a comprehensive F1 score of 76.24% on the Tusimple and CULane datasets, respectively. Real vehicle tests show that the algorithm detects a frame rate of 67 fps, allowing for realtime detection in complex traffic scenarios and more effectively addressing the problem of lane line occlusion.

, 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=Junzhao JIANG, Bin PENG, Wenhao YANG, Yekai XU), CN=ArticleExt(id=1153809152473289161, articleId=1153809125017375052, tenantId=1146029695717560320, journalId=1152916057816748034, language=CN, title=基于注意力机制与线锚信息传递的车道线检测, columnId=1153756967408488571, journalTitle=汽车工程学报, columnName=智能网联技术专栏/主编:高镇海, runingTitle=null, highlight=null, articleAbstract=

车道线检测是自动驾驶领域的关键技术,目前仍面临较多挑战。车道线监督信号的稀疏性以及复杂场景下的遮挡、阴影等因素会影响检测的准确率与实时性。基于此,提出了一种融合CBAM注意力机制与线锚特征聚合模块的车道线检测模型,提出的算法在Tusimple 和 CULane数据集分别达到96.19%的准确率和76.24%的综合F1得分,通过实车测试表明,该算法检测帧率为67fps,可以在复杂交通场景下进行实时检测,较好地解决了车道线遮挡问题。

, correspAuthors=null, authorNote=null, correspAuthorsNote=

彭彬(1996-),男,河南驻马店人,硕士研究生,主要研究方向为深度学习车道线检测。Tel:18437903286 E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=zW0Hwh4x+d6TodG4De0tOQ==, magXml=wxnUh3aFRmtljPPsd2GUrA==, pdfUrl=null, pdf=NsB33/UWNIYkk/A9+1YJlw==, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=RSCYSnQ3XmWC60+SWaUTqg==, mapNumber=null, authorCompany=null, fund=null, authors=

姜俊昭(1987-),男,山东烟台人,博士,讲师,主要研究方向为高级别自动驾驶的感知融合算法以及基于数据驱动的整车测评技术。Tel:13637051538 E-mail:

, authorsList=姜俊昭, 彭彬, 杨文豪, 徐业凯)}, authors=[Author(id=1153809153538642394, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=chlgch.2006@hfut.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1153809153601556957, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, authorId=1153809153538642394, language=EN, stringName=Junzhao JIANG, firstName=Junzhao, middleName=null, lastName=JIANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 Hefei University of Technology Hefei 230041 China
2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1153809153656082910, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, authorId=1153809153538642394, language=CN, stringName=姜俊昭, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 合肥工业大学 合肥 230041
2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041, bio={"img":"45e3eOVYlRTn36sK3NJpMA==","content":"

姜俊昭(1987-),男,山东烟台人,博士,讲师,主要研究方向为高级别自动驾驶的感知融合算法以及基于数据驱动的整车测评技术。Tel:13637051538 E-mail:

"}, bioImg=45e3eOVYlRTn36sK3NJpMA==, bioContent=

姜俊昭(1987-),男,山东烟台人,博士,讲师,主要研究方向为高级别自动驾驶的感知融合算法以及基于数据驱动的整车测评技术。Tel:13637051538 E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1153809153396036051, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=1, ext=[AuthorCompanyExt(id=1153809153400230356, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Hefei University of Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153404424661, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 合肥工业大学 合肥 230041)]), AuthorCompany(id=1153809153454756310, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=2, ext=[AuthorCompanyExt(id=1153809153458950615, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153467339224, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041)])]), Author(id=1153809153718997472, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=pengbin7238@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1153809153819660771, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, authorId=1153809153718997472, language=EN, stringName=Bin PENG, firstName=Bin, middleName=null, lastName=PENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 Hefei University of Technology Hefei 230041 China
2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1153809153874186724, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, authorId=1153809153718997472, language=CN, stringName=彭彬, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 合肥工业大学 合肥 230041
2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1153809153396036051, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=1, ext=[AuthorCompanyExt(id=1153809153400230356, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Hefei University of Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153404424661, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 合肥工业大学 合肥 230041)]), AuthorCompany(id=1153809153454756310, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=2, ext=[AuthorCompanyExt(id=1153809153458950615, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153467339224, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041)])]), Author(id=1153809153937101286, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, 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=1153809154000015849, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, authorId=1153809153937101286, language=EN, stringName=Wenhao YANG, firstName=Wenhao, middleName=null, lastName=YANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 Hefei University of Technology Hefei 230041 China
2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1153809154075513323, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, authorId=1153809153937101286, language=CN, stringName=杨文豪, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 合肥工业大学 合肥 230041
2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1153809153396036051, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=1, ext=[AuthorCompanyExt(id=1153809153400230356, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Hefei University of Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153404424661, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 合肥工业大学 合肥 230041)]), AuthorCompany(id=1153809153454756310, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=2, ext=[AuthorCompanyExt(id=1153809153458950615, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153467339224, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041)])]), Author(id=1153809154130039278, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, 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=1153809154192953843, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, authorId=1153809154130039278, language=EN, stringName=Yekai XU, firstName=Yekai, middleName=null, lastName=XU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 Hefei University of Technology Hefei 230041 China
2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1153809154272645624, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, authorId=1153809154130039278, language=CN, stringName=徐业凯, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1 合肥工业大学 合肥 230041
2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1153809153396036051, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=1, ext=[AuthorCompanyExt(id=1153809153400230356, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Hefei University of Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153404424661, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 合肥工业大学 合肥 230041)]), AuthorCompany(id=1153809153454756310, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=2, ext=[AuthorCompanyExt(id=1153809153458950615, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153467339224, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041)])])], keywords=[Keyword(id=1153809154721436166, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, orderNo=1, keyword=lane line detection), Keyword(id=1153809154775962121, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, orderNo=2, keyword=line anchors), Keyword(id=1153809154843070988, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, orderNo=3, keyword=attention mechanisms), Keyword(id=1153809154889208334, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, orderNo=4, keyword=information delivery), Keyword(id=1153809154939539985, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, orderNo=1, keyword=车道线检测), Keyword(id=1153809155015037461, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, orderNo=2, keyword=线锚), Keyword(id=1153809155094729239, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, orderNo=3, keyword=注意力机制), Keyword(id=1153809155166032410, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, orderNo=4, keyword=信息传递)], refs=[Reference(id=1153809157984604824, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2005, volume=null, issue=null, pageStart=706, pageEnd=711, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=CHIU KY, LIN S F, journalName=Proceedings of the 2005 IEEE Intelligent Vehicles Symposium, June 6-8, 2005, Las Vegas, NV, USA.Piscataway NJ: IEEE, refType=null, unstructuredReference=CHIU KY, LIN S F. Lane Detection Using Color-Based Segmentation[C]// Proceedings of the 2005 IEEE Intelligent Vehicles Symposium, June 6-8, 2005, Las Vegas, NV, USA.Piscataway NJ: IEEE, 2005:706-711., articleTitle=Lane Detection Using Color-Based Segmentation, refAbstract=null), Reference(id=1153809158051713691, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2006, volume=null, issue=null, pageStart=1168, pageEnd=1172, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=SUNTY, TSAI S J, CHAN V, journalName=Proceedings of the 2006 IEEE Intelligent Transportation Systems Conference, Sept. 17-20, 2006, Toronto, ON, Canada. Piscataway NJ: IEEE, refType=null, unstructuredReference=SUNTY, TSAI S J, CHAN V. HSI Color Model Based Lane-Marking Detection[C]// Proceedings of the 2006 IEEE Intelligent Transportation Systems Conference, Sept. 17-20, 2006, Toronto, ON, Canada. Piscataway NJ: IEEE, 2006:1168-1172., articleTitle=HSI Color Model Based Lane-Marking Detection, refAbstract=null), Reference(id=1153809158123016861, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2014, volume=25, issue=3, pageStart=727, pageEnd=745, url=null, language=null, rfNumber=[3], rfOrder=2, authorNames=HILLEL A B, LERNER R, DAN L, journalName=Machine Vision & Applications, refType=null, unstructuredReference=HILLEL A B, LERNER R, DAN L, et al. Recent Progress in Road and Lane Detection: A Survey[J]. Machine Vision & Applications, 2014,25(3):727-745., articleTitle=Recent Progress in Road and Lane Detection: A Survey, refAbstract=null), Reference(id=1153809158265623199, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2012, volume=17, issue=6, pageStart=657, pageEnd=663, url=null, language=null, rfNumber=[4], rfOrder=3, authorNames=王永忠, 王晓云, 文成林, journalName=中国图象图形学报, refType=null, unstructuredReference=王永忠, 王晓云, 文成林. 梯度点对约束的结构化车道检测[J]. 中国图象图形学报, 2012,17(6):657-663., articleTitle=梯度点对约束的结构化车道检测, refAbstract=null), Reference(id=1153809158374675105, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2012, volume=17, issue=6, pageStart=657, pageEnd=663, url=null, language=null, rfNumber=[4], rfOrder=4, authorNames=WANG Yongzhong, WANG Xiaoyun, WEN Chenglin, journalName=Journal of Image and Graphics, refType=null, unstructuredReference=WANG Yongzhong, WANG Xiaoyun, WEN Chenglin. Gradient-Pair Constraint for Structure Lane Detection[J]. Journal of Image and Graphics, 2012,17(6):657-663. (in Chinese), articleTitle=Gradient-Pair Constraint for Structure Lane Detection, refAbstract=null), Reference(id=1153809158433395363, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2013, volume=17, issue=2, pageStart=188, pageEnd=199, url=null, language=null, rfNumber=[5], rfOrder=5, authorNames=YOU Feng, ZHANG Ronghui, ZHONG Lingshu, journalName=Journal of the Optical Society of Korea, refType=null, unstructuredReference=YOU Feng, ZHANG Ronghui, ZHONG Lingshu, et al. Lane Detection Algorithm for Night-Time Digital Image Based on Distribution Feature of Boundary Pixels[J]. Journal of the Optical Society of Korea, 2013,17(2):188-199., articleTitle=Lane Detection Algorithm for Night-Time Digital Image Based on Distribution Feature of Boundary Pixels, refAbstract=null), Reference(id=1153809158492115621, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2017, volume=39, issue=3, pageStart=357, pageEnd=363, url=null, language=null, rfNumber=[6], rfOrder=6, authorNames=陈无畏, 蒋玉亭, 谈东奎, journalName=汽车工程, refType=null, unstructuredReference=陈无畏, 蒋玉亭, 谈东奎. 一种基于边缘点投影的车道线快速识别算法[J]. 汽车工程, 2017,39(3):357-363., articleTitle=一种基于边缘点投影的车道线快速识别算法, refAbstract=null), Reference(id=1153809158538252966, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2017, volume=39, issue=3, pageStart=357, pageEnd=363, url=null, language=null, rfNumber=[6], rfOrder=7, authorNames=CHEN Wuwei, JIANG Yuting, TAN Dongkui, journalName=Automotive Engineering, refType=null, unstructuredReference=CHEN Wuwei, JIANG Yuting, TAN Dongkui. A Fast Lane Marking Recognition Algorithm Based on Edge Projection[J]. Automotive Engineering, 2017,39(3):357-363. (in Chinese), articleTitle=A Fast Lane Marking Recognition Algorithm Based on Edge Projection, refAbstract=null), Reference(id=1153809158584390311, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2019, volume=9, issue=1, pageStart=27, pageEnd=35, url=null, language=null, rfNumber=[7], rfOrder=8, authorNames=骆济焕, 兰凤崇, 陈吉清, journalName=汽车工程学报, refType=null, unstructuredReference=骆济焕, 兰凤崇, 陈吉清. 联合图像处理和目标约束的车道线检测方法[J]. 汽车工程学报, 2019,9(1):27-35., articleTitle=联合图像处理和目标约束的车道线检测方法, refAbstract=null), Reference(id=1153809158634721960, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2019, volume=9, issue=1, pageStart=27, pageEnd=35, url=null, language=null, rfNumber=[7], rfOrder=9, authorNames=LUO Jihuan, LAN Fengchong, CHEN Jiqing, journalName=Chinese Journal of Automotive Engineering, refType=null, unstructuredReference=LUO Jihuan, LAN Fengchong, CHEN Jiqing. Lane Line Detection Method Based on Joint Image Processing and Target Constraint[J]. Chinese Journal of Automotive Engineering, 2019,9(1):27-35. (in Chinese), articleTitle=Lane Line Detection Method Based on Joint Image Processing and Target Constraint, refAbstract=null), Reference(id=1153809158693442217, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2013, volume=null, issue=null, pageStart=59, pageEnd=64, url=null, language=null, rfNumber=[8], rfOrder=10, authorNames=LIU Guorong, LI Shutao, LIU Weirong, journalName=Proceedings of the 2013 IEEE Chinese Automation Congress, Nov. 7-8, 2013, Changsha, China. Piscataway NJ: IEEE, refType=null, unstructuredReference=LIU Guorong, LI Shutao, LIU Weirong. Lane Detection Algorithm Based on Local Feature Extraction[C]// Proceedings of the 2013 IEEE Chinese Automation Congress, Nov. 7-8, 2013, Changsha, China. Piscataway NJ: IEEE, 2013:59-64., articleTitle=Lane Detection Algorithm Based on Local Feature Extraction, refAbstract=null), Reference(id=1153809158760551083, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2018, volume=8, issue=3, pageStart=189, pageEnd=195, url=null, language=null, rfNumber=[9], rfOrder=11, authorNames=时利, 谈东奎, 王乃汉, journalName=汽车工程学报, refType=null, unstructuredReference=时利, 谈东奎, 王乃汉, 等. 一种基于特征点投票统计的车道线快速检测算法[J]. 汽车工程学报, 2018,8(3):189-195., articleTitle=一种基于特征点投票统计的车道线快速检测算法, refAbstract=null), Reference(id=1153809158810882733, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2018, volume=8, issue=3, pageStart=189, pageEnd=195, url=null, language=null, rfNumber=[9], rfOrder=12, authorNames=SHI Li, TAN Dongkui, WANG Naihan, journalName=Chinese Journal of Automotive Engineering, refType=null, unstructuredReference=SHI Li, TAN Dongkui, WANG Naihan, et al. A Rapid Lane Line Detection Algorithm Based on Feature Point Voting Statistics[J]. Chinese Journal of Automotive Engineering, 2018,8(3):189-195. (in Chinese), articleTitle=A Rapid Lane Line Detection Algorithm Based on Feature Point Voting Statistics, refAbstract=null), Reference(id=1153809158865408687, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2014, volume=null, issue=null, pageStart=454, pageEnd=461, url=null, language=null, rfNumber=[10], rfOrder=13, authorNames=KIM JH, LEE M H, journalName=Proceedings of the 2014 International Conference on Neural Information Processing, Nov. 3-6, 2014, Kuching, Malaysia. ICNIP, refType=null, unstructuredReference=KIM JH, LEE M H. Robust Lane Detection Based on Convolutional Neural Network and Random Sample Consensus[C]// Proceedings of the 2014 International Conference on Neural Information Processing, Nov. 3-6, 2014, Kuching, Malaysia. ICNIP, 2014:454-461., articleTitle=Robust Lane Detection Based on Convolutional Neural Network and Random Sample Consensus, refAbstract=null), Reference(id=1153809158957683382, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2017, volume=28, issue=3, pageStart=690, pageEnd=703, url=null, language=null, rfNumber=[11], rfOrder=14, authorNames=LI Jun, MEI Xue, PROKHOROV D, journalName=IEEE Transactions on Neural Networks and Learning Systems, refType=null, unstructuredReference=LI Jun, MEI Xue, PROKHOROV D, et al. Deep Neural Network for Structural Prediction and Lane Detection in Traffic Scene[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017,28(3):690-703., articleTitle=Deep Neural Network for Structural Prediction and Lane Detection in Traffic Scene, refAbstract=null), Reference(id=1153809159037375161, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=1013, pageEnd=1021, url=null, language=null, rfNumber=[12], rfOrder=15, authorNames=HOU Yuenan, MA Zheng, LIU Chunxiao, journalName=Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision, Oct. 27-Nov. 2, 2019, Seoul, Korea (South). Piscataway NJ: IEEE, refType=null, unstructuredReference=HOU Yuenan, MA Zheng, LIU Chunxiao, et al. Learning Lightweight Lane Detection CNNs by Self Attention Distillation[C]// Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision, Oct. 27-Nov. 2, 2019, Seoul, Korea (South). Piscataway NJ: IEEE, 2019:1013-1021., articleTitle=Learning Lightweight Lane Detection CNNs by Self Attention Distillation, refAbstract=null), Reference(id=1153809159091901115, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=7276, pageEnd=7283, url=null, language=null, rfNumber=[13], rfOrder=16, authorNames=PAN Xingang, SHI Jianping, LUO Ping, journalName=Proceedings of the 2018 AAAI Conference on Artificial Intelligence, New Orleans.AAAI, refType=null, unstructuredReference=PAN Xingang, SHI Jianping, LUO Ping, et al. Spatial as Deep: Spatial CNN for Traffic Scene Understanding[C]// Proceedings of the 2018 AAAI Conference on Artificial Intelligence, New Orleans.AAAI, 2018:7276-7283., articleTitle=Spatial as Deep: Spatial CNN for Traffic Scene Understanding, refAbstract=null), Reference(id=1153809159154815677, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=276, pageEnd=291, url=null, language=null, rfNumber=[14], rfOrder=17, authorNames=QIN Zequn, WANG Huanyu, LI Xi, journalName=Proceedings of the 2020 European Conference on Computer Vision, Aug. 23-28, Glasgow, UK.ECCV, refType=null, unstructuredReference=QIN Zequn, WANG Huanyu, LI Xi. Ultra Fast Structure-Aware Deep Lane Detection[C]// Proceedings of the 2020 European Conference on Computer Vision, Aug. 23-28, Glasgow, UK.ECCV, 2020:276-291., articleTitle=Ultra Fast Structure-Aware Deep Lane Detection, refAbstract=null), Reference(id=1153809159217730241, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=286, pageEnd=291, url=null, language=null, rfNumber=[15], rfOrder=18, authorNames=NEVEN D, DE BRABANDERE B, GEORGOULIS S, journalName=Proceedings of the 2018 IEEE Intelligent Vehicles Symposium (IV), June 26-30, 2018, Changshu, China. Piscataway NJ: IEEE, refType=null, unstructuredReference=NEVEN D, DE BRABANDERE B, GEORGOULIS S, et al. Towards End-to-End Lane Detection: An Instance Segmentation Approach[C]// Proceedings of the 2018 IEEE Intelligent Vehicles Symposium (IV), June 26-30, 2018, Changshu, China. Piscataway NJ: IEEE, 2018:286-291., articleTitle=Towards End-to-End Lane Detection: An Instance Segmentation Approach, refAbstract=null), Reference(id=1153809159289033412, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=6150, pageEnd=6156, url=null, language=null, rfNumber=[16], rfOrder=19, authorNames=TABELINI L, BERRIEL R, PAIXAO T M, journalName=Proceedings of the 2021 IEEE International Conference on Pattern Recognition, Jan. 10-15, 2021, Milan, Italy. Piscataway NJ: IEEE, refType=null, unstructuredReference=TABELINI L, BERRIEL R, PAIXAO T M, et al. PolyLaneNet: Lane Estimation via Deep Polynomial Regression[C]// Proceedings of the 2021 IEEE International Conference on Pattern Recognition, Jan. 10-15, 2021, Milan, Italy. Piscataway NJ: IEEE, 2021:6150-6156., articleTitle=PolyLaneNet: Lane Estimation via Deep Polynomial Regression, refAbstract=null), Reference(id=1153809159347753671, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=1006, pageEnd=1007, url=null, language=null, rfNumber=[17], rfOrder=20, authorNames=YOO SW, LEE HS, MYEONG H, journalName=Proceedings of the 2020 IEEE Conference on Computer Vision and Pattern Recognition, June 14-19, 2020, Seattle, USA.Piscataway NJ: IEEE, refType=null, unstructuredReference=YOO SW, LEE HS, MYEONG H, et al. End-to-End Lane Marker Detection via Row-Wise Classification[C]// Proceedings of the 2020 IEEE Conference on Computer Vision and Pattern Recognition, June 14-19, 2020, Seattle, USA.Piscataway NJ: IEEE, 2020:1006-1007., articleTitle=End-to-End Lane Marker Detection via Row-Wise Classification, refAbstract=null), Reference(id=1153809159435834057, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=1947, pageEnd=1955, url=null, language=null, rfNumber=[18], rfOrder=21, authorNames=LEE S, KIM J, SHIN YOON J, journalName=Proceedings of the 2017 IEEE International Conference on Computer Vision, Oct. 22-29, 2017, Venice, Italy. Piscataway NJ: IEEE, refType=null, unstructuredReference=LEE S, KIM J, SHIN YOON J, et al. VPGNet: Vanishing Point Guided Network for Lane and Road Marking Detection and Recognition[C]// Proceedings of the 2017 IEEE International Conference on Computer Vision, Oct. 22-29, 2017, Venice, Italy. Piscataway NJ: IEEE, 2017:1947-1955., articleTitle=VPGNet: Vanishing Point Guided Network for Lane and Road Marking Detection and Recognition, refAbstract=null), Reference(id=1153809159507137228, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2023, volume=45, issue=2, pageStart=263, pageEnd=272, url=null, language=null, rfNumber=[19], rfOrder=22, authorNames=武志斐, 李守彪, journalName=汽车工程, refType=null, unstructuredReference=武志斐, 李守彪. 基于实例分割的车道线检测算法[J]. 汽车工程, 2023,45(2):263-272., articleTitle=基于实例分割的车道线检测算法, refAbstract=null), Reference(id=1153809159599411919, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2023, volume=45, issue=2, pageStart=263, pageEnd=272, url=null, language=null, rfNumber=[19], rfOrder=23, authorNames=WU Zhifei, LI Shoubiao, journalName=Automotive Engineering, refType=null, unstructuredReference=WU Zhifei, LI Shoubiao. Lane Line Detection Algorithm Based on Instance Segmentation[J]. Automotive Engineering, 2023,45(2):263-272. (in Chinese), articleTitle=Lane Line Detection Algorithm Based on Instance Segmentation, refAbstract=null), Reference(id=1153809159662326483, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=689, pageEnd=704, url=null, language=null, rfNumber=[20], rfOrder=24, authorNames=XU Hang, WANG Shaoju, CAI Xinyue, journalName=Proceedings of the 2020 Computer Vision-ECCV, Aug. 23-28, 2020, Glasgow, UK.ECCV, refType=null, unstructuredReference=XU Hang, WANG Shaoju, CAI Xinyue, et al. CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending[C]// Proceedings of the 2020 Computer Vision-ECCV, Aug. 23-28, 2020, Glasgow, UK.ECCV, 2020:689-704., articleTitle=CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending, refAbstract=null), Reference(id=1153809159716852436, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2020, volume=21, issue=1, pageStart=248, pageEnd=258, url=null, language=null, rfNumber=[21], rfOrder=25, authorNames=LI Xiang, LI Jun, HU Xiaolin, journalName=IEEE Transactions on Intelligent Transportation Systems, refType=null, unstructuredReference=LI Xiang, LI Jun, HU Xiaolin, et al. Line-CNN: End-to-End Traffic Line Detection with Line Proposal Unit[J]. IEEE Transactions on Intelligent Transportation Systems, 2020,21(1):248-258., articleTitle=Line-CNN: End-to-End Traffic Line Detection with Line Proposal Unit, refAbstract=null), Reference(id=1153809159775572694, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=3694, pageEnd=3702, url=null, language=null, rfNumber=[22], rfOrder=26, authorNames=LIU Ruijin, YUAN Zejian, LIU Tie, journalName=Proceedings of the 2021 IEEE/CVF Winter Conference on Applications of Computer Vision, Jan. 5-9, 2021, Waikoloa, HI, USA. Piscataway NJ: IEEE, refType=null, unstructuredReference=LIU Ruijin, YUAN Zejian, LIU Tie, et al. End-to-End Lane Shape Prediction with Transformers[C]// Proceedings of the 2021 IEEE/CVF Winter Conference on Applications of Computer Vision, Jan. 5-9, 2021, Waikoloa, HI, USA. Piscataway NJ: IEEE, 2021:3694-3702., articleTitle=End-to-End Lane Shape Prediction with Transformers, refAbstract=null), Reference(id=1153809159838487257, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[23], rfOrder=27, authorNames=VASWANI A, SHAZEER N M, PARMAR N, journalName=Conference and Workshop on Neural Information Processing Systems (NIPS), refType=null, unstructuredReference=VASWANI A, SHAZEER N M, PARMAR N, et al. "Attention is All you Need"[C]// Conference and Workshop on Neural Information Processing Systems (NIPS), 2017., articleTitle="Attention is All you Need", refAbstract=null), Reference(id=1153809159897207513, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=294, pageEnd=302, url=null, language=null, rfNumber=[24], rfOrder=28, authorNames=TABELINI L, BERRIEL R, PAIXAO T M, journalName=Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 19-25, 2021, Nashville, TN, USA. Piscataway NJ: IEEE, refType=null, unstructuredReference=TABELINI L, BERRIEL R, PAIXAO T M, et al. Keep Your Eyes on the Lane: Real-Time Attention-Guided Lane Detection[C]// Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 19-25, 2021, Nashville, TN, USA. Piscataway NJ: IEEE, 2021:294-302., articleTitle=Keep Your Eyes on the Lane: Real-Time Attention-Guided Lane Detection, refAbstract=null), Reference(id=1153809159964316380, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=3, pageEnd=19, url=null, language=null, rfNumber=[25], rfOrder=29, authorNames=WOO S, PARK J, LEE JY, journalName=Proceedings of the 2018 European Conference on Computer Vision, Sept. 8-14, 2018, Munich, Germany.ECCV, refType=null, unstructuredReference=WOO S, PARK J, LEE JY, et al. Cbam: Convolutional Block Attention Module[C]// Proceedings of the 2018 European Conference on Computer Vision, Sept. 8-14, 2018, Munich, Germany.ECCV, 2018:3-19., articleTitle=Cbam: Convolutional Block Attention Module, refAbstract=null), Reference(id=1153809160027230943, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=770, pageEnd=778, url=null, language=null, rfNumber=[26], rfOrder=30, authorNames=HE Kaiming, ZHANG Xiangyu, REN Shaoqing, journalName=Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, June 27-30, 2016, Las Vegas, NV, USA.Piscataway NJ: IEEE, refType=null, unstructuredReference=HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Deep Residual Learning for Image Recognition[C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, June 27-30, 2016, Las Vegas, NV, USA.Piscataway NJ: IEEE, 2016:770-778., articleTitle=Deep Residual Learning for Image Recognition, refAbstract=null), Reference(id=1153809160081756896, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=3547, pageEnd=3554, url=null, language=null, rfNumber=[27], rfOrder=31, authorNames=ZHENG Tu, FANG Hao, ZHANG Yi, journalName=Proceedings of the 2021 AAAI Conference on Artificial Intelligence, Feb. 2-9, 2021, USA.AAAI, refType=null, unstructuredReference=ZHENG Tu, FANG Hao, ZHANG Yi, et al. Resa: Recurrent Feature-Shift Aggregator for Lane Detection[C]// Proceedings of the 2021 AAAI Conference on Artificial Intelligence, Feb. 2-9, 2021, USA.AAAI, 2021:3547-3554., articleTitle=Resa: Recurrent Feature-Shift Aggregator for Lane Detection, refAbstract=null), Reference(id=1153809160153060066, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, doi=null, pmid=null, pmcid=null, year=2019, volume=5, issue=4, pageStart=417, pageEnd=428, url=null, language=null, rfNumber=[28], rfOrder=32, authorNames=FAN Ruochen, WANG Xuanrun, HOU Qibin, journalName=Computational Visual Media, refType=null, unstructuredReference=FAN Ruochen, WANG Xuanrun, HOU Qibin, et al. SpinNet: Spinning Convolutional Network for Lane Boundary Detection[J]. Computational Visual Media, 2019,5(4):417-428., articleTitle=SpinNet: Spinning Convolutional Network for Lane Boundary Detection, refAbstract=null)], funds=[Fund(id=1153809157779083921, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, awardId=JZ2021AFKJ0002, language=CN, fundingSource=安徽省新能源汽车暨智能网联汽车创新工程项目(JZ2021AFKJ0002), fundOrder=null, country=null), Fund(id=1153809157846192787, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, awardId=JZ2023YQTD0073, language=CN, fundingSource=中央高校基本科研业务费专项资金资助项目(JZ2023YQTD0073), fundOrder=null, country=null), Fund(id=1153809157900718741, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, awardId=PA2023GDSK0113, language=CN, fundingSource=安徽省重点实验室自主创新专项(PA2023GDSK0113), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1153809153396036051, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=1, ext=[AuthorCompanyExt(id=1153809153400230356, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Hefei University of Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153404424661, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153396036051, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 合肥工业大学 合肥 230041)]), AuthorCompany(id=1153809153454756310, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, xref=2, ext=[AuthorCompanyExt(id=1153809153458950615, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China), AuthorCompanyExt(id=1153809153467339224, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, companyId=1153809153454756310, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041)])], figs=[ArticleFig(id=1153809156046836279, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=HFyNh6u2jTZ6BYQ8bCt39w==, figureFileBig=BUEiAKyYH15loyLLXlLHMQ==, tableContent=null), ArticleFig(id=1153809156113945145, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=图 1, caption=模型设计, figureFileSmall=HFyNh6u2jTZ6BYQ8bCt39w==, figureFileBig=BUEiAKyYH15loyLLXlLHMQ==, tableContent=null), ArticleFig(id=1153809156164276795, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=XqO1Icf8YZxSGRl5tRCxqw==, figureFileBig=K/kpGka1nytZONIGn38lkw==, tableContent=null), ArticleFig(id=1153809156210414141, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=图 2, caption=$\mathrm{{CBAM}}$ 注意力模块, figureFileSmall=XqO1Icf8YZxSGRl5tRCxqw==, figureFileBig=K/kpGka1nytZONIGn38lkw==, tableContent=null), ArticleFig(id=1153809156256551487, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=syXJl1cuRbsOPnViqwnvxA==, figureFileBig=Fcof3zFX9v6yPnv8WMCzkg==, tableContent=null), ArticleFig(id=1153809156311077441, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=图 4, caption=空间注意模块, figureFileSmall=syXJl1cuRbsOPnViqwnvxA==, figureFileBig=Fcof3zFX9v6yPnv8WMCzkg==, tableContent=null), ArticleFig(id=1153809156365603398, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=mj9SnyPVfeTxUOQS6SGzBw==, figureFileBig=iaHOpygz/Aq83bCvH27SHg==, tableContent=null), ArticleFig(id=1153809156436906571, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=图 5, caption=线锚信息传递模块, figureFileSmall=mj9SnyPVfeTxUOQS6SGzBw==, figureFileBig=iaHOpygz/Aq83bCvH27SHg==, tableContent=null), ArticleFig(id=1153809156520792653, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=ABxfXCa7EKpORuOr1792aQ==, figureFileBig=ufzlQ42pySJeEi2/XV1MJQ==, tableContent=null), ArticleFig(id=1153809156583707216, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=图 6, caption=测试图片, figureFileSmall=ABxfXCa7EKpORuOr1792aQ==, figureFileBig=ufzlQ42pySJeEi2/XV1MJQ==, tableContent=null), ArticleFig(id=1153809156642427475, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=2RrCMR6CsJNVtCwcPn99LQ==, figureFileBig=Uhujtk28u+WaAAdujuiqxQ==, tableContent=null), ArticleFig(id=1153809156688564821, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=图 7, caption=实车测试效果图, figureFileSmall=2RrCMR6CsJNVtCwcPn99LQ==, figureFileBig=Uhujtk28u+WaAAdujuiqxQ==, tableContent=null), ArticleFig(id=1153809156734702167, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
参数 值/方式
训练次数 100; 200
batch_size 8
optimizer Adam
Learning rate 0.000 3
Input size ${360} \times {640}$
学习率衰减方式 CosineAnnealingLR
$N$ 72
${\delta }_{\text{back }}$ 32
${N}_{\text{anc }}$ 1000
$K$ 9
$\xi$ 15
$\mu$ 20
$\lambda$ 10
$\omega$ 9
$r$ 16
), ArticleFig(id=1153809156793422425, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=表 1, caption=模型训练超参数设置, figureFileSmall=null, figureFileBig=null, tableContent=
参数 值/方式
训练次数 100; 200
batch_size 8
optimizer Adam
Learning rate 0.000 3
Input size ${360} \times {640}$
学习率衰减方式 CosineAnnealingLR
$N$ 72
${\delta }_{\text{back }}$ 32
${N}_{\text{anc }}$ 1000
$K$ 9
$\xi$ 15
$\mu$ 20
$\lambda$ 10
$\omega$ 9
$r$ 16
), ArticleFig(id=1153809156843754076, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Acc/% fp fn
$\omega = 1$ 95.88 0.034 5 0.0123
$\omega = 3$ 95.83 0.0139 0.0158
$\omega = 5$ 95.73 0.0157 0.0168
$\omega = 7$ 95.92 0.0176 0.0116
$\omega = 9$ 96.19 0.025 5 0.0144
$\omega = {11}$ 95.82 0.014 7 0.0125
), ArticleFig(id=1153809156894085729, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=表 2, caption=在 TuSimple 上不同卷积核宽度对准确率的影响, figureFileSmall=null, figureFileBig=null, tableContent=
Acc/% fp fn
$\omega = 1$ 95.88 0.034 5 0.0123
$\omega = 3$ 95.83 0.0139 0.0158
$\omega = 5$ 95.73 0.0157 0.0168
$\omega = 7$ 95.92 0.0176 0.0116
$\omega = 9$ 96.19 0.025 5 0.0144
$\omega = {11}$ 95.82 0.014 7 0.0125
), ArticleFig(id=1153809156944417382, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
F1/% Precision Recall
$\omega = 1$ 76.08 0.814 3 0.714 0
$\omega = 3$ 75.72 0.805 3 0.714 6
$\omega = 5$ 75.40 0.821 8 0.696 6
$\omega = 7$ 75.09 0.8170 0.694 8
$\omega = 9$ 76.24 0.805 4 0.7238
$\omega = {11}$ 75.34 0.8023 0.7101
), ArticleFig(id=1153809157015720554, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=表 3, caption=在 CULane 上不同卷积核宽度对 F1 的影响, figureFileSmall=null, figureFileBig=null, tableContent=
F1/% Precision Recall
$\omega = 1$ 76.08 0.814 3 0.714 0
$\omega = 3$ 75.72 0.805 3 0.714 6
$\omega = 5$ 75.40 0.821 8 0.696 6
$\omega = 7$ 75.09 0.8170 0.694 8
$\omega = 9$ 76.24 0.805 4 0.7238
$\omega = {11}$ 75.34 0.8023 0.7101
), ArticleFig(id=1153809157074440813, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
ACC/% fp fn
Resnet34 94.72 0.051 4 0.0184
Resnet34+CBAM 95.89 0.031 1 0.0149
Resnet34+信息传递模块 95.93 0.0136 0.0114
Resnet34+CBAM+信息传递模块 96.19 0.025 5 0.014 4
), ArticleFig(id=1153809157137355378, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=表 4, caption=TuSimple 上加入不同模块的试验对比, figureFileSmall=null, figureFileBig=null, tableContent=
ACC/% fp fn
Resnet34 94.72 0.051 4 0.0184
Resnet34+CBAM 95.89 0.031 1 0.0149
Resnet34+信息传递模块 95.93 0.0136 0.0114
Resnet34+CBAM+信息传递模块 96.19 0.025 5 0.014 4
), ArticleFig(id=1153809157221241461, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
Precision Recall
Resnet34 75.34 0.807 1 0.706 5
Resnet34+CBAM 75.59 0.802 2 0.714 6
Resnet34+信息传递模块 75.98 0.832 5 0.698 9
Resnet34+CBAM+信息传递模块 76.24 0.805 4 0.7238
), ArticleFig(id=1153809157279961721, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=表 5, caption=CULane 上加入不同模块的试验对比, figureFileSmall=null, figureFileBig=null, tableContent=
Precision Recall
Resnet34 75.34 0.807 1 0.706 5
Resnet34+CBAM 75.59 0.802 2 0.714 6
Resnet34+信息传递模块 75.98 0.832 5 0.698 9
Resnet34+CBAM+信息传递模块 76.24 0.805 4 0.7238
), ArticleFig(id=1153809157334487678, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
场景 SCNN SpinNet [ 28 ] 实例分割 [ 19 ] 本文模型
Total 71.63 74.20 75.20 76.24
Normal 90.60 90.50 91.10 91.09
Crowded 69.70 71.70 73.10 72.51
Night 66.10 68.10 70.50 68.37
No line 43.40 43.20 46.60 47.13
Shadow 66.90 72.90 74.10 67.70
Arrow 84.10 85.00 85.80 86.60
Hight 58.50 62.00 67.10 64.43
Curve 64.40 50.70 69.80 67.29
Cross 1990 1923 1 298
FPS 8.30 69.20 70.10 113
), ArticleFig(id=1153809157439345282, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=表 6, caption=不同网络对比试验, figureFileSmall=null, figureFileBig=null, tableContent=
场景 SCNN SpinNet [ 28 ] 实例分割 [ 19 ] 本文模型
Total 71.63 74.20 75.20 76.24
Normal 90.60 90.50 91.10 91.09
Crowded 69.70 71.70 73.10 72.51
Night 66.10 68.10 70.50 68.37
No line 43.40 43.20 46.60 47.13
Shadow 66.90 72.90 74.10 67.70
Arrow 84.10 85.00 85.80 86.60
Hight 58.50 62.00 67.10 64.43
Curve 64.40 50.70 69.80 67.29
Cross 1990 1923 1 298
FPS 8.30 69.20 70.10 113
), ArticleFig(id=1153809157548397192, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
设备 配置参数
Nuvo-5095GC 工控机 GPU: Independent NVIDIA® GPU (75W TDP)
CPU: Intel® Core™ i5-6500TE
Ubuntu版本: 18.04
工作温度:-25~60 °C
LI-USB30-AR023ZWDRB 摄像机 分辨率: ${1920} \times {1080}$ 工作电压: +5VDC 帧率: 30 fps
), ArticleFig(id=1153809157640671884, tenantId=1146029695717560320, journalId=1152916057816748034, articleId=1153809125017375052, language=CN, label=表 7, caption=设备及参数, figureFileSmall=null, figureFileBig=null, tableContent=
设备 配置参数
Nuvo-5095GC 工控机 GPU: Independent NVIDIA® GPU (75W TDP)
CPU: Intel® Core™ i5-6500TE
Ubuntu版本: 18.04
工作温度:-25~60 °C
LI-USB30-AR023ZWDRB 摄像机 分辨率: ${1920} \times {1080}$ 工作电压: +5VDC 帧率: 30 fps
)], attaches=null, journal=Journal(id=1149663730306166798, delFlag=0, nameCn=汽车工程学报, nameEn=Chinese Journal of Automotive Engineering, nameHistory1=null, nameHistory2=null, issn=2095-1469, eissn=null, cn=50-1206/U, coden=null, periodic=1, language=CN, oaType=null, 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=jNBfZbtw0oWUAJwGjDXBiw==, journalPrice=null, startedYear=null, abbrevIsoEn=, journalRemark=null, publicationField=null, createdTime=1752030793026, updatedTime=1761730209297, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=Q, firstLetterEn=Q, subjectCode=Engineering, subjectName=工程, subjectCodeEn=Engineering, subjectNameEn=null, picCn=jNBfZbtw0oWUAJwGjDXBiw==, picEn=AzxyR8a1Jh9nr4is1i+eyQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1190346030903296709, language=CN, name=汽车工程学报, 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=, createdTime=1761730209327, updatedTime=1761730209327, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://tougao.ijournals.cn/ch/index.aspx, submissionEditorUrl=http://tougao.ijournals.cn/ch/index.aspx, submissionReviewUrl=http://tougao.ijournals.cn/ch/index.aspx, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1190346030945239750, language=EN, name=Chinese Journal of Automotive Engineering, 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=1761730209337, updatedTime=1761730209337, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://tougao.ijournals.cn/ch/index.aspx, submissionEditorUrl=http://tougao.ijournals.cn/ch/index.aspx, submissionReviewUrl=http://tougao.ijournals.cn/ch/index.aspx, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1152916057816748034, websiteList=[Website(id=1153018156433002908, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1152916057816748034, 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/qcgcxb/CN, language=CN, createTime=1752830550535, createBy=18614031015, updateTime=1752831104570, updateBy=18614031015, name=汽车工程学报, tplId=1146099689490845704, title=汽车工程学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1154109936599552608, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156433002908, code=articleTextType, value=kx, createTime=1753090851211, updateTime=1753090851211, creator=18614031015, updator=18614031015), WebsiteProps(id=1154109936578581085, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156433002908, code=banner, value=null, createTime=1753090851206, updateTime=1753090851206, creator=18614031015, updator=18614031015), WebsiteProps(id=1154109936561803868, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156433002908, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=AAkd3aJi43D2QHSdjNHtbQ==, createTime=1753090851202, updateTime=1753090851202, creator=18614031015, updator=18614031015), WebsiteProps(id=1154109936591163999, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156433002908, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1753090851209, updateTime=1753090851209, creator=18614031015, updator=18614031015), WebsiteProps(id=1154109936586969694, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156433002908, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1753090851208, updateTime=1753090851208, creator=18614031015, updator=18614031015)]), Website(id=1153018156516888991, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1152916057816748034, 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/qcgcxb/EN, language=EN, createTime=1752830550556, createBy=18614031015, updateTime=1752831098136, updateBy=18614031015, name=汽车工程学报, tplId=1146101810881728533, title=汽车工程学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1154109989737189989, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156516888991, code=articleTextType, value=hep, createTime=1753090863880, updateTime=1753090863880, creator=18614031015, updator=18614031015), WebsiteProps(id=1154109989716218466, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156516888991, code=banner, value=null, createTime=1753090863875, updateTime=1753090863875, creator=18614031015, updator=18614031015), WebsiteProps(id=1154109989695246945, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156516888991, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/EN/file/pic?fileId=AAkd3aJi43D2QHSdjNHtbQ==, createTime=1753090863870, updateTime=1753090863870, creator=18614031015, updator=18614031015), WebsiteProps(id=1154109989732995684, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156516888991, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/EN/file/pic, createTime=1753090863879, updateTime=1753090863879, creator=18614031015, updator=18614031015), WebsiteProps(id=1154109989724607075, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1153018156516888991, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1753090863877, updateTime=1753090863877, creator=18614031015, updator=18614031015)])], journalTitle=汽车工程学报, weixinUrl=null, journalUrl=null, iacademicId=null, status=1, seqNo=null, journalTitleEn=Chinese Journal of Automotive Engineering, journalPhotoCn=jNBfZbtw0oWUAJwGjDXBiw==, journalPhotoEn=AzxyR8a1Jh9nr4is1i+eyQ==, journalFirstLetter=Q, 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/qcgcxb/CN/10.3969/j.issn.2095–1469.2024.05.07, detailUrlEn=https://castjournals.cast.org.cn/joweb/qcgcxb/EN/10.3969/j.issn.2095–1469.2024.05.07, pdfUrlCn=https://castjournals.cast.org.cn/joweb/qcgcxb/CN/PDF/10.3969/j.issn.2095–1469.2024.05.07, pdfUrlEn=https://castjournals.cast.org.cn/joweb/qcgcxb/EN/PDF/10.3969/j.issn.2095–1469.2024.05.07, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于注意力机制与线锚信息传递的车道线检测
收藏切换
PDF下载
姜俊昭 1, 2 , 彭彬 1, 2 , 杨文豪 1, 2 , 徐业凯 1, 2
汽车工程学报 | 智能网联技术专栏/主编:高镇海 2024,14(5): 812-820
收起
收藏切换
汽车工程学报 | 智能网联技术专栏/主编:高镇海 2024, 14(5): 812-820
基于注意力机制与线锚信息传递的车道线检测
全屏
姜俊昭1, 2 , 彭彬1, 2 , 杨文豪1, 2, 徐业凯1, 2
作者信息
  • 1 合肥工业大学 合肥 230041
  • 2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041
  • 姜俊昭(1987-),男,山东烟台人,博士,讲师,主要研究方向为高级别自动驾驶的感知融合算法以及基于数据驱动的整车测评技术。Tel:13637051538 E-mail:

通讯作者:


彭彬(1996-),男,河南驻马店人,硕士研究生,主要研究方向为深度学习车道线检测。Tel:18437903286 E-mail:
Lane Line Detection Based on Attention Mechanism and Line Anchor Information Transmission
Junzhao JIANG1, 2 , Bin PENG1, 2 , Wenhao YANG1, 2, Yekai XU1, 2
Affiliations
  • 1 Hefei University of Technology Hefei 230041 China
  • 2 Key Laboratory for Automated Vehicle Safety Technology Hefei 230041 China
doi: 10.3969/j.issn.2095–1469.2024.05.07
文章导航
收藏切换

车道线检测是自动驾驶领域的关键技术,目前仍面临较多挑战。车道线监督信号的稀疏性以及复杂场景下的遮挡、阴影等因素会影响检测的准确率与实时性。基于此,提出了一种融合CBAM注意力机制与线锚特征聚合模块的车道线检测模型,提出的算法在Tusimple 和 CULane数据集分别达到96.19%的准确率和76.24%的综合F1得分,通过实车测试表明,该算法检测帧率为67fps,可以在复杂交通场景下进行实时检测,较好地解决了车道线遮挡问题。

车道线检测  /  线锚  /  注意力机制  /  信息传递

Lane line detection is a key technology in the field of autonomous driving, and it currently faces many challenges. The sparsity of the lane line supervision signal, as well as factors such as occlusion and shadows in complex scenes, can affect detection accuracy and realtime performance. Based on this, this paper proposes a lane line detection model that integrates the CBAM attention mechanism and a line anchor feature aggregation module. The proposed algorithm achieves an accuracy of 96.19% and a comprehensive F1 score of 76.24% on the Tusimple and CULane datasets, respectively. Real vehicle tests show that the algorithm detects a frame rate of 67 fps, allowing for realtime detection in complex traffic scenarios and more effectively addressing the problem of lane line occlusion.

lane line detection  /  line anchors  /  attention mechanisms  /  information delivery
姜俊昭, 彭彬, 杨文豪, 徐业凯. 基于注意力机制与线锚信息传递的车道线检测. 汽车工程学报, 2024 , 14 (5) : 812 -820 . DOI: 10.3969/j.issn.2095–1469.2024.05.07
Junzhao JIANG, Bin PENG, Wenhao YANG, Yekai XU. Lane Line Detection Based on Attention Mechanism and Line Anchor Information Transmission[J]. Chinese Journal of Automotive Engineering, 2024 , 14 (5) : 812 -820 . DOI: 10.3969/j.issn.2095–1469.2024.05.07
相较于常规的目标检测或语义分割, 车道线检测通常更具有挑战性, 因为车道线的外观为细长条状, 特征单一且极易被遮挡, 车道线的这种特点给其检测带来了误检和漏检的风险。对于车道线检测,有两种主流方法,即传统图像处理方法 [ 1 - 9 ] 和深度学习方法 [ 10 - 18 ] 。传统车道线检测方法主要通过手工提取不同场景下的车道线颜色 [ 1 - 3 ] 、梯度变化 [ 4 - 5 ] 或边缘特征 [ 6 - 9 ] ,采用霍夫变化、最小二乘法等后处理方式进行车道线的形状拟合。当实际道路环境中车道线不被车辆遮挡且较清晰时, 传统图像处理方法可以表现出良好的准确性和检测效率, 但是在光线变化、车道线模糊、车道线被遮挡及夜间低照度等复杂场景下却难以取得理想效果, 很难满足真实路况场景下对算法的鲁棒性要求。近年来, 由于具有良好的表示和学习能力, 深度学习方法在该领域取得了巨大的成功, 但仍有一些重要且具有挑战性的问题需要解决。车道线检测算法作为自动驾驶的核心组成部分, 为下游的规划和决策模块提供支撑, 这需要较低的车道线检测计算成本和较好的鲁棒性。
根据对车道线的不同理解, 当前基于深度学习的车道线检测主要可分为基于分割的算法 [ 12 - 13 , 15 , 19 ] 、 基于锚的算法 [ 14 , 21 ] 和基于模型参数 [ 16 , 22 ] 的算法。 在深度学习取代传统的基于手工选择特征的车道线检测方法以后, 逐像素预测车道线的方式被广泛采用。LaneNet融合了二值语义分割与聚类实例分割, 突破了以往仅限于固定数量的车道线检测, 缺点是后处理中的聚类步骤很耗时, 很难满足实时性要求 [ 15 ] 。SCNN在分割任务中沿不同维度方向使用片卷积来从不同维度聚合车道线像素信息, 虽然片卷积模块极大提升了检测精度, 但是却牺牲了效率 [ 13 ] 。SAD 提出了一种自注意提取模块来聚合上下文信息 [ 12 ] 。该模块允许使用更轻的主干网络, 在效率和性能方面达到了一个很好的平衡。 VPGNet 提出了一种由车道线消失点引导的多任务网络, 将车道线大多在远处聚集为一点作为先验信息, 用于车道线和道路标记检测, 其精度和鲁棒性都达到了较好的效果,且能实时运行 [ 18 ] 。 CurveLanes NAS 使用神经架构搜索来找到更好的主干,尽管取得了先进的成果, 但其 NAS 在计算上非常昂贵 [ 20 ] 。逐像素预测方式虽然精度较高,但往往预测很多不必要的像素点, 且聚类和曲线拟合等后处理方式降低了车道线检测速度。基于线锚的车道线检测方法起源于目标检测中的锚框, 和锚框不同的是, 线锚是从具有特定方向的图像边界发射的直线, 通过对直线的优化来获得驾驶场景中车道线的真实形状。Line-CNN 首次提出基于线建议来预测车道线, 为车道线检测提供了新的思路, 虽然该方式达到较高的检测精度, 但是在模型检测速度方面表现一般 [ 21 ] 。UFLD 提出一种基于行方向上的位置选择的车道线检测方式, 显著降低计算成本, 达到很高的检测速度, 但该方式是以牺牲检测精度为代价 [ 14 ] 。PolyLaneNet 提出了基于深度多项式回归的模型, 在该方法中, 模型学习为每条车道线输出多项式, 尽管检测速度很快, 但其预测结果却显示了对直线车道线的高度偏向 [ 16 ] 。LSTR [ 22 ] 将 Transformer [ 23 ] 引入车道线检测,用来捕获车道线细长的结构和全局上下文, 以最轻巧的模型尺寸和最快的速度显示了最新的准确性, 但是存在模型曲线方程参数难以优化的问题。
基于上述分析, 各种车道线检测方法各有其特点, 综合考虑算法实车部署时对算力的限制, 本文采用基于线锚 [ 24 ] 的方式来预测车道线,将车道线建模为一系列的坐标集合, 避免了密集预测带来的检测效率不高等问题, CBAM注意力模块 [ 25 ] 和线锚信息传递模块在其中起到重要作用。注意力模块可以帮助网络关注重要的车道线特征并抑制不必要的特征。线锚信息传递模块则用于汇聚线锚全局信息,这在解决遮挡、车道线磨损等问题时是有必要的。
本文所提的基于注意力机制与特征聚合的车道线检测是用于车道线检测的基于线锚的单阶段模型, 该方法的模型设计如 图 1 所示。它接收来自摄像头拍摄的包含车道线的图像, 输出为车道线拟合曲线。为了实现这一过程, 输入图像首先经过神经网络骨架进行初步特征提取, 由于车道线的细长形特征, 本文在神经网络骨架后又加入通道和空间注意力模块, 使网络自身更加关注车道线细节信息。 在经过若干预定义的线锚对特征图中的车道线信息进行采样后, 又添加了线锚信息传递模块以汇聚线锚全局信息, 通过结合局部信息和全局信息, 该模型可以更容易使用来自其他车道线的信息。
车道线由一系列的二维坐标来表示,其中 $Y$ 坐标由一系列的等距点 ${\left\{ {y}_{i}\right\} }_{i = 0}^{N - 1}$ 表示,因为 $Y$ 被固定, 所以车道线可以由 $X = {\left\{ {x}_{i}\right\} }_{i = 0}^{N - 1}$ 唯一确定。由于大多数车道线不垂直穿过整个图像, 所以使用起始索引 sts 和结束索引 eds 来定义 $X$ 的有效连续序列。所采用的方法采用线的方式执行基于锚的检测, 因此, 车道建议以一系列射线来作为参考。线锚由除图像顶部的 3 个边界上的起始点 $O$ 和方向 $\alpha$ 定义。当图像上锚的数量足够多时, 便可表示绝大多数真实车道线。
网络的第一阶段是特征提取, 本文采用 ResNet34 [ 26 ] 作为网络骨架,用来提取初步的图像特征,生成的特征图 ${F}^{\prime } \in {C}_{F} \times {H}_{F} \times {W}_{F}$ ,然后,再将特征图送至 CBAM 注意力模块, 获得优化后的特征图 $F \in {C}_{F} \times {H}_{F} \times {W}_{F}$ , CBAM注意力模块如 图 2 所示。
该模块有两个顺序的子模块: 通道注意模块 ( 图 3 ) 和空间注意模块 ( 图 4 )。
通道注意模块由沿空间方向的平均池化和最大池化并联组成, 平均池化和最大池化操作聚合特征图的空间信息, 生成两个不同的空间上下文描述符 ${F}_{\mathrm{{avg}}}^{\mathrm{c}}$${F}_{\max }^{\mathrm{c}}$ ,它们分别表示平均池化特征和最大池化特征, 然后将两个不同的空间上下文描述符转发至多层感知机 (MLP), 为了减少参数, 多层感知机中间层被设置为 ${R}^{\frac{C}{r} \times 1 \times 1}$ ,在将多层感知机应用于每个描述符之后, 使用逐元素求和合并输出特征向量, 再将特征向量经过 sigmoid 函数激活获得通道注意 ${M}_{a} \in {R}^{C \times 1 \times 1}$ ,计算过程如下。
$ {M}_{a}\left( {F}^{\prime }\right) = \operatorname{MLP}\left( {\operatorname{AvgPool}\left( {F}^{\prime }\right) }\right) + \\ \operatorname{MLP}\left( {\operatorname{MaxPool}\left( {F}^{\prime }\right) }\right) \text{。} \\ {F}^{\prime \prime } = {M}_{a}\left( {F}^{\prime }\right) \otimes {F}^{\prime }\text{。} $
0197d13f-f604-751b-82b8-d18b9a721fe0_2_959_240_557_164_0.jpg
图3 通道注意模块
式中: $\operatorname{AvgPool}\left( {F}^{\prime }\right)$ 为特征 ${F}^{\prime }$ 的平均池化; $\operatorname{MaxPool}\left( {F}^{\prime }\right)$ 为特征 ${F}^{\prime }$ 的最大池化。AvgPool $\left( {F}^{\prime }\right)$ 和MaxPool $\left( {F\text{’}}\right)$ 作为中间输入共享 MLP 权重,经通道注意优化后的特征为 ${F}^{\prime \prime }$
空间注意模块位于通道注意力模块之后, 由沿通道方向的平均池化和最大池化并联组成, 生成两个二维图 ${F}_{\text{avg }}^{\mathrm{c}} \in {R}^{1 \times H \times W}$${F}_{\max }^{\mathrm{c}} \in {R}^{1 \times H \times W}$ ,将 ${F}_{\text{avg }}^{\mathrm{c}}$${F}_{\max }^{\mathrm{c}}$ 串联起来以生成有效的特征描述符,然后通过标准卷积层将它们连接起来并卷积, 再经过 sigmoid 函数激活生成 $2\mathrm{D}$ 空间注意图,计算过程如下。
$ {M}_{s}\left( {F}^{\prime \prime }\right) = \left( {{\mathcal{F}}^{7 \times 7}\left( \left\lbrack {\operatorname{AvgPool}\left( {F}^{\prime \prime }\right) ;\operatorname{MaxPool}\left( {F}^{\prime \prime }\right) }\right\rbrack \right) }\right) $
$ F = {M}_{s}\left( {F}^{\prime \prime }\right) \otimes {F}^{\prime \prime }\text{。} $
式中: ${\mathcal{F}}^{7 \times 7}$ 为滤波器大小为 $7 \times 7$ 的卷积运算; ${F}^{\prime \prime }$ 为经通道注意模块处理后的特征; $\operatorname{AvgPool}\left( {F}^{\prime \prime }\right)$ 为特征 ${F}^{\prime }$ 的平均池化; MaxPool $\left( {F}^{\prime \prime }\right)$ 为特征 ${F}^{\prime \prime }$ 的最大池化,最终获得特征 $F$
线锚定义了特征图 $F$ 上的点,这些点被用于对应的车道建议,而线锚由一系列的坐标点 ${\left\{ \left( {x}_{i},{y}_{i}\right) \right\} }_{i = 0}^{N - 1}$ 表示,由 1.1 节所述,给定 ${y}_{i} =$ $0,1,2,\ldots , N - 1$ ,则相应的 ${x}_{i}$ 可以计算为:
$ {x}_{i} = \left\lfloor {\frac{1}{\tan \alpha }\left( {{y}_{i} - \frac{{y}_{o}}{{\delta }_{\text{back }}}}\right) + \frac{{x}_{o}}{{\delta }_{\text{back }}}}\right\rfloor 。 $
式中: $\left( {{x}_{o},{y}_{o}}\right)$$\alpha$ 分别为线锚的原点和斜率; $\left( {{x}_{i},{y}_{i}}\right)$ 为线锚上的坐标; ${\delta }_{\text{back }}$ 为主干网络的全局步幅。
为了提取特征图 $\mathrm{F}$ 上的车道线信息,每个线锚 $i$ 将由对应的特征向量 ${a}_{i}^{\text{loc}} \in {R}^{{C}_{F} \cdot {H}_{F}}$ 来表示,该特征向量从携带车道线局部信息的 $\mathrm{F}$ 集合而来。如果线锚的一部分位于 $F$ 边界之外的情况下,则该部分 ${a}_{i}^{\text{loc }}$ 使用 0 来填充,如果预定义的线锚个数为 ${N}_{\mathrm{{anc}}}$ ,则 ${A}^{\mathrm{{loc}}} = {\left\lbrack \begin{array}{llll} {a}_{0}^{\mathrm{{loc}}}, & {a}_{1}^{\mathrm{{loc}}}, & \ldots , & {a}_{{N}_{\mathrm{{anc}}} - 1}^{\mathrm{{loc}}} \end{array}\right\rbrack }^{\mathrm{T}}$ ,其中 ${A}^{\mathrm{{loc}}} \in {R}^{{N}_{\mathrm{{anc}}} \times {C}_{F} \cdot {H}_{F}}$
线锚作为车道线坐标回归的基准, 承载着车道线的坐标信息, 根据其定义, 不同车道线由不同线锚作为基准进行优化, 而不同线锚之间并没有信息的交互,这决定了某一车道线可能不能很好地利用来自其他车道线的信息, 对于车道线被遮挡的情况, 可能难以预测车道线。因此, 如何将不同线锚之间的信息进行传递, 可能会影响到困难场景下的算法表现,根据线锚的这种特点,本文对 RESA [ 27 ] 的特征聚合模块进行改进, 并将其加入到不同线锚之间, 使局部的线锚能接收到来自全局的信息, 线锚信息传递模块的定义为:
$ {G}_{v, p, i}^{k} = \mathop{\sum }\limits_{{a, b}}{D}_{a, c, b} \cdot {X}_{a,\left( {p + {s}_{k}}\right) {\;\operatorname{mod}\;H}, i + b - 1}^{k} \circ $
$ {X}_{v, p, i}^{{k}^{\prime }} = {X}_{v, p, i,}^{k} + f\left( {G}_{v, p, i}^{k}\right) 。 $
$ {s}_{k} = \frac{L}{{2}^{K - k}}, k = 0,1,\cdots , K - 1。 $
${A}^{\text{loc }} \in {R}^{{N}_{\text{anc }} \times {C}_{F} \cdot {H}_{F}}$ 扩充为 ${A}_{\varphi }^{\text{loc }} \in {R}^{1 \times {N}_{\text{anc }} \times {C}_{F} \cdot {H}_{F}}$ 来匹配线锚信息传递模块中所需的三维输入, 其中, 1、 ${N}_{\text{anc }}$${C}_{F} \cdot {H}_{F}$ 分别表示通道、线锚总数和线锚总长; ${X}_{v, p, i}^{k}$ 表示第 $k$ 次迭代时特征 ${A}_{\varphi }^{\mathrm{{loc}}}$ 的值,其中, $v\text{、}p$$i$ 分别表示通道、线锚和线锚位置的索引; $K = \left| {{\log }_{2}L}\right| ;k$ 为迭代次数, $L$ 在式 (8) 中为 ${C}_{F} \cdot {H}_{F}$$f$ 为非线性激活函数;上标为 ’ 的 $X$ 表示已更新的元素; ${s}_{k}$ 为第 $k$ 次迭代中的移位步长。式 (6)为垂直特征聚合公式, $D$ 为一组一维卷积核, 其大小为 ${N}_{\text{in }} \times {N}_{\text{out }} \times \omega$ ,其中, ${N}_{\text{in }}\text{、}{N}_{\text{out }}$$\omega$ 分别表示输入通道的数量、输出通道的数量和核宽度, ${N}_{\text{in }}$${N}_{\text{out }}$ 都等于 1,式 (6) 中的 $G$ 是特征聚合的中间结果。
图 5 显示了当步长为 1 和 2 时线锚信息传递模块的信息传递,其中,切片高度为 ${N}_{\mathrm{{anc}}}$ ,宽度为 ${C}_{F} \cdot {H}_{F}$ ,沿竖直方向共 $s$ 个切片,其中 $s = \left\lfloor \frac{{N}_{\text{anc }}}{\omega }\right\rfloor$
基于车道的线锚来预测车道建议, 其由以下 3 个部分组成:
1)$K + 1$ 概率,其中 $K$ 表示车道类别数;
2)$N$ 个偏移量(预测点和线锚点之间的水平距离);
3)车道建议的长度 $l$ (有效偏移的数量)。
车道建议的起始索引 sts 直接由线锚原点的 ${y}_{o}$ 坐标确定。因此,结束索引可确定为 eds = sts+ $\left\lbrack l\right\rbrack - 1$ 。经过线锚信息传递模块处理后的特征会经过两个全连接层来预测最终的车道建议, 一个全连接层用于执行分类,分类的概率 ${p}_{i} =$ $\left\{ {{p}_{0},\ldots ,{p}_{K + 1}}\right\}$ ,另一个全连接层用于执行回归, 回归的数值 ${r}_{i} = \left\{ {l,\left\{ {{x}_{0},\ldots ,{x}_{N - 1}}\right\} }\right\}$
非极大值抑制用于测量任意一条线锚和任意一条车道线之间的距离, 当预测的车道建议和真值车道线距离低于设定阈值 $\xi$ 时被分配一个正的标签, 而预测的车道建议和真值车道线距离高于设定阈值 $\mu$ 时被分配一个负的标签。
在模型训练期间, 经非极大值抑制过滤后的既没有被分配正标签也没有被分配负标签的锚将会被舍弃, 相应地, 保留下来的锚将会用于焦点损失函数和平滑 ${L}_{1}$ 损失函数,其中,焦点损失函数用于处理分类目标,平滑 ${L}_{1}$ 损失函数用于处理回归目标。
锚的数量是限制模型速度的主要因素之一, 锚的数量越多,模型运行的速度就越慢。本文保留了在数据集中出现次数较多的锚(数量为 ${N}_{\text{anc }}$ ),使模型在训练和验证期间的运行速度都有所提升。
本文分别在车道线检测数据集 TuSimple 和 CULane 上进行了试验测试, 定性地评估本文方法的有效性和效率。在下面 8 个小节中, 本文将分别介绍 TuSimple 和 CULane 数据集构成、实施细节、 评价标准及试验结果。
TuSimple 作为 CVPR 2017 自动驾驶挑战研讨会正式比赛的一部分被发布。该数据集的主要特点是, 图像是在高速公路上相对理想的天气和照明条件下拍摄的, 数据集中包含的最大车道线数量为 5 , 其中, 3268 张图片被用于训练, 358 张图片被用于验证, 2 782 张图片被用于测试, 所有图像都有 ${1280} \times {720}$ 像素。它是车道线检测工作中使用最广泛的数据集之一。
CULane 是用于车道检测学术研究的大规模具有挑战性的数据集。它由安装在 6 辆不同驾驶员驾驶的汽车上的摄像机收集, 收集了超过 55 小时的视频, 并提取了 133235 帧。数据集被分为 88880 个训练集, 9 675 个验证集和 34680 个测试集, 所有图像都有 ${1640} \times {590}$ 像素。其中,测试集分为正常和 8 个挑战性类别。
本文算法基于Ubuntu18.04系统与PyTorch1.10.0 深度学习框架进行训练及测试。试验硬件环境为搭载 Intel Xeon(R)Gold 6226R CPU @2.90GHz $\times {64}$ 处理器, 64 G 内存, 同时利用 NVIDIA RTX A6000 48G双显卡并行计算的高性能工作站。
网络超参数见 表 1 , 在训练次数上, 由于数据集之间的差异, 在 CULane 数据集上的训练次数设置为 200,在 TuSimple 上的训练次数设置为 100。
对于 TuSimple, 本文用准确度指标来衡量, 准确度的定义如式(9)所示。
$ \mathrm{{Acc}} = \frac{\mathop{\sum }\limits_{\text{clip }}{C}_{\text{clip }}}{\mathop{\sum }\limits_{\text{clip }}{S}_{\text{clip }}}。 $
式中: ${C}_{\text{clip }}$ 为输出图像中正确预测的车道线点的数量; ${S}_{\text{clip }}$ 为输出图像中总的车道线点的数量,只有当预测点在真值车道的 20 个像素内时, 预测点才被认为是正确预测。
对于 CULane,本文用 F1 指标来衡量, F1 指标的定义为:
$ \text{F1} - \text{Score} = 2 \times \frac{\text{Presision} \times \text{Recall}}{\text{Presision} + \text{Recall}}\text{。} $
$ \text{Presision} = \frac{\mathrm{{TP}}}{\mathrm{{TP}} + \mathrm{{FP}}}\text{。} $
$ \text{Recall} = \frac{\mathrm{{TP}}}{\mathrm{{TP}} + \mathrm{{FN}}}\text{。} $
TP 表示预测像素被判定为正样本, 实际上也是正样本, FP 表示预测像素被判定为正样本, 实际上是负样本, FN 表示预测像素被判定为负样本, 实际上是正样本。
还有一个重要评价指标为 FPS, 其表征模型预测图像的速度,定义如式(13)所示。
$ \text{ FPS } = \frac{1000}{s}\text{ 。 } $
式中: $s$ 为单帧图像的处理时间,单位ms。
在线锚信息传递模块中, 不同线锚之间的信息传递是通过一维卷积的移动迭代来实现的, 信息传递的效果可能受卷积核宽度大小的影响, 表 2表 3 显示了不同卷积核宽度对最终结果的影响。
表 2表 3 分别展示了该网络在 tusimple 和 culane 数据集上不同卷积核大小下的验证准确率, 当卷积核宽度 $\omega = 9$ 时,在 tusimple上的准确率和在 culane上的 F1 得分达到最高。
为验证本网络所采用模块的有效性, 本文在试验参数保持不变的情况下做了如下对比,试验结果见 表 4表 5
表 4表 5 分别展示了在基线网络上加入不同模块的试验效果变化,其中 $\omega = 9$ 。试验结果表明, 在网络中加入 CBAM 注意力模块和线锚信息传递模块后, 在 TuSimple 上的准确率和在 CULane 上的 F1 得分分别提升了 1.6% 和 1.12%,验证了所使用模块在基于线锚的车道线检测方法中的有效性。
表 6 展示了本文所提网络模型和其他网络模型在 CULane 上的对比试验。
表 6 可知, 本文模型在道路箭头、无车道线和综合场景下的 F1 得分达到最优, 在其他场景下也取得较好效果, 且本文模型的网络实际运行速度达到了 SCNN 的 13 倍以上, 验证了所提模型在车道线遮挡和综合场景下的有效性。
图 6 显示了该网络在 TuSimple 和 CULane 上测试的实际效果, 其中, 第 1 行为 TuSimple 部分图片的测试效果, 第 2 行为 CULane 部分图片的测试效果。 图 6 中, 蓝色线为车道线真值标注, 绿色线为预测出的车道线。在本文的试验中, 尽管部分车道线被车辆严重遮挡,网络仍能较好地预测出车道线。
为了验证本文算法的实用性, 将该算法部署到车载工控机的 ROS 平台上进行验证。本测试采用的车载工控机型号为 Nuvo-5095GC,摄像机型号为 LI-USB30-AR023ZWDRB, 详细参数见 表 7
首先将单目摄像机安装在车辆挡风玻璃后侧中上位置, 并进行内参和外参的标定。为验证该网络在实际道路场景下的实时性和准确性, 选取上海市环城高速作为本次测试的地点, 将车辆行驶速度调整为 ${70}\mathrm{\;{km}}/\mathrm{h}$ 进行测试,结果表明,本文算法实际测试时检测每帧图像耗时约 ${67}\mathrm{\;{ms}}$ ,满足实时性的要求。使用rviz可视化工具查看车道线分割效果, 检测结果如 图 7 所示。
在实际道路测试时, 所提出的方法能根据道路车辆及环境检测出被遮挡车道线。
本文采用基于线锚的方法来预测车道线, 充分利用车道线先验知识, 避免了密集预测导致的检测速度慢等问题, 同时, 在网络中采用不同方式来关注车道细节、解决车道线遮挡问题。经验证, 该网络能在 TuSimple 数据集上实现 96.19% 的预测准确率, 在 CULane 数据集上实现 76.24% 的 F1 得分且运行速度能达到 SCNN 的 13 倍以上。最后,通过实车测试,该网络检测每帧图像耗时约 ${67}\mathrm{\;{ms}}$ 。结果表明, 该网络满足实时性的要求, 并对遮挡场景有较高的泛化性和鲁棒性。
  • 安徽省新能源汽车暨智能网联汽车创新工程项目(JZ2021AFKJ0002)
  • 中央高校基本科研业务费专项资金资助项目(JZ2023YQTD0073)
  • 安徽省重点实验室自主创新专项(PA2023GDSK0113)
参考文献 引证文献
排序方式:
[1]
CHIU KY, LIN S F. Lane Detection Using Color-Based Segmentation[C]// Proceedings of the 2005 IEEE Intelligent Vehicles Symposium, June 6-8, 2005, Las Vegas, NV, USA.Piscataway NJ: IEEE, 2005:706-711.
[2]
SUNTY, TSAI S J, CHAN V. HSI Color Model Based Lane-Marking Detection[C]// Proceedings of the 2006 IEEE Intelligent Transportation Systems Conference, Sept. 17-20, 2006, Toronto, ON, Canada. Piscataway NJ: IEEE, 2006:1168-1172.
[3]
HILLEL A B, LERNER R, DAN L, et al. Recent Progress in Road and Lane Detection: A Survey[J]. Machine Vision & Applications, 2014,25(3):727-745.
[4]
王永忠, 王晓云, 文成林. 梯度点对约束的结构化车道检测[J]. 中国图象图形学报, 2012,17(6):657-663.
WANG Yongzhong, WANG Xiaoyun, WEN Chenglin. Gradient-Pair Constraint for Structure Lane Detection[J]. Journal of Image and Graphics, 2012,17(6):657-663. (in Chinese)
[5]
YOU Feng, ZHANG Ronghui, ZHONG Lingshu, et al. Lane Detection Algorithm for Night-Time Digital Image Based on Distribution Feature of Boundary Pixels[J]. Journal of the Optical Society of Korea, 2013,17(2):188-199.
[6]
陈无畏, 蒋玉亭, 谈东奎. 一种基于边缘点投影的车道线快速识别算法[J]. 汽车工程, 2017,39(3):357-363.
CHEN Wuwei, JIANG Yuting, TAN Dongkui. A Fast Lane Marking Recognition Algorithm Based on Edge Projection[J]. Automotive Engineering, 2017,39(3):357-363. (in Chinese)
[7]
骆济焕, 兰凤崇, 陈吉清. 联合图像处理和目标约束的车道线检测方法[J]. 汽车工程学报, 2019,9(1):27-35.
LUO Jihuan, LAN Fengchong, CHEN Jiqing. Lane Line Detection Method Based on Joint Image Processing and Target Constraint[J]. Chinese Journal of Automotive Engineering, 2019,9(1):27-35. (in Chinese)
[8]
LIU Guorong, LI Shutao, LIU Weirong. Lane Detection Algorithm Based on Local Feature Extraction[C]// Proceedings of the 2013 IEEE Chinese Automation Congress, Nov. 7-8, 2013, Changsha, China. Piscataway NJ: IEEE, 2013:59-64.
[9]
时利, 谈东奎, 王乃汉, 等. 一种基于特征点投票统计的车道线快速检测算法[J]. 汽车工程学报, 2018,8(3):189-195.
SHI Li, TAN Dongkui, WANG Naihan, et al. A Rapid Lane Line Detection Algorithm Based on Feature Point Voting Statistics[J]. Chinese Journal of Automotive Engineering, 2018,8(3):189-195. (in Chinese)
[10]
KIM JH, LEE M H. Robust Lane Detection Based on Convolutional Neural Network and Random Sample Consensus[C]// Proceedings of the 2014 International Conference on Neural Information Processing, Nov. 3-6, 2014, Kuching, Malaysia. ICNIP, 2014:454-461.
[11]
LI Jun, MEI Xue, PROKHOROV D, et al. Deep Neural Network for Structural Prediction and Lane Detection in Traffic Scene[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017,28(3):690-703.
[12]
HOU Yuenan, MA Zheng, LIU Chunxiao, et al. Learning Lightweight Lane Detection CNNs by Self Attention Distillation[C]// Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision, Oct. 27-Nov. 2, 2019, Seoul, Korea (South). Piscataway NJ: IEEE, 2019:1013-1021.
[13]
PAN Xingang, SHI Jianping, LUO Ping, et al. Spatial as Deep: Spatial CNN for Traffic Scene Understanding[C]// Proceedings of the 2018 AAAI Conference on Artificial Intelligence, New Orleans.AAAI, 2018:7276-7283.
[14]
QIN Zequn, WANG Huanyu, LI Xi. Ultra Fast Structure-Aware Deep Lane Detection[C]// Proceedings of the 2020 European Conference on Computer Vision, Aug. 23-28, Glasgow, UK.ECCV, 2020:276-291.
[15]
NEVEN D, DE BRABANDERE B, GEORGOULIS S, et al. Towards End-to-End Lane Detection: An Instance Segmentation Approach[C]// Proceedings of the 2018 IEEE Intelligent Vehicles Symposium (IV), June 26-30, 2018, Changshu, China. Piscataway NJ: IEEE, 2018:286-291.
[16]
TABELINI L, BERRIEL R, PAIXAO T M, et al. PolyLaneNet: Lane Estimation via Deep Polynomial Regression[C]// Proceedings of the 2021 IEEE International Conference on Pattern Recognition, Jan. 10-15, 2021, Milan, Italy. Piscataway NJ: IEEE, 2021:6150-6156.
[17]
YOO SW, LEE HS, MYEONG H, et al. End-to-End Lane Marker Detection via Row-Wise Classification[C]// Proceedings of the 2020 IEEE Conference on Computer Vision and Pattern Recognition, June 14-19, 2020, Seattle, USA.Piscataway NJ: IEEE, 2020:1006-1007.
[18]
LEE S, KIM J, SHIN YOON J, et al. VPGNet: Vanishing Point Guided Network for Lane and Road Marking Detection and Recognition[C]// Proceedings of the 2017 IEEE International Conference on Computer Vision, Oct. 22-29, 2017, Venice, Italy. Piscataway NJ: IEEE, 2017:1947-1955.
[19]
武志斐, 李守彪. 基于实例分割的车道线检测算法[J]. 汽车工程, 2023,45(2):263-272.
WU Zhifei, LI Shoubiao. Lane Line Detection Algorithm Based on Instance Segmentation[J]. Automotive Engineering, 2023,45(2):263-272. (in Chinese)
[20]
XU Hang, WANG Shaoju, CAI Xinyue, et al. CurveLane-NAS: Unifying Lane-Sensitive Architecture Search and Adaptive Point Blending[C]// Proceedings of the 2020 Computer Vision-ECCV, Aug. 23-28, 2020, Glasgow, UK.ECCV, 2020:689-704.
[21]
LI Xiang, LI Jun, HU Xiaolin, et al. Line-CNN: End-to-End Traffic Line Detection with Line Proposal Unit[J]. IEEE Transactions on Intelligent Transportation Systems, 2020,21(1):248-258.
[22]
LIU Ruijin, YUAN Zejian, LIU Tie, et al. End-to-End Lane Shape Prediction with Transformers[C]// Proceedings of the 2021 IEEE/CVF Winter Conference on Applications of Computer Vision, Jan. 5-9, 2021, Waikoloa, HI, USA. Piscataway NJ: IEEE, 2021:3694-3702.
[23]
VASWANI A, SHAZEER N M, PARMAR N, et al. "Attention is All you Need"[C]// Conference and Workshop on Neural Information Processing Systems (NIPS), 2017.
[24]
TABELINI L, BERRIEL R, PAIXAO T M, et al. Keep Your Eyes on the Lane: Real-Time Attention-Guided Lane Detection[C]// Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 19-25, 2021, Nashville, TN, USA. Piscataway NJ: IEEE, 2021:294-302.
[25]
WOO S, PARK J, LEE JY, et al. Cbam: Convolutional Block Attention Module[C]// Proceedings of the 2018 European Conference on Computer Vision, Sept. 8-14, 2018, Munich, Germany.ECCV, 2018:3-19.
[26]
HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Deep Residual Learning for Image Recognition[C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, June 27-30, 2016, Las Vegas, NV, USA.Piscataway NJ: IEEE, 2016:770-778.
[27]
ZHENG Tu, FANG Hao, ZHANG Yi, et al. Resa: Recurrent Feature-Shift Aggregator for Lane Detection[C]// Proceedings of the 2021 AAAI Conference on Artificial Intelligence, Feb. 2-9, 2021, USA.AAAI, 2021:3547-3554.
[28]
FAN Ruochen, WANG Xuanrun, HOU Qibin, et al. SpinNet: Spinning Convolutional Network for Lane Boundary Detection[J]. Computational Visual Media, 2019,5(4):417-428.
2024年第14卷第5期
PDF下载
274
109
引用本文
BibTeX
文章信息
doi: 10.3969/j.issn.2095–1469.2024.05.07
  • 接收时间:2023-07-31
  • 首发时间:2025-07-20
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2023-07-31
  • 修回日期:2023-09-15
基金
安徽省新能源汽车暨智能网联汽车创新工程项目(JZ2021AFKJ0002)
中央高校基本科研业务费专项资金资助项目(JZ2023YQTD0073)
安徽省重点实验室自主创新专项(PA2023GDSK0113)
作者信息
    1 合肥工业大学 合肥 230041
    2 安徽省自动驾驶汽车安全技术重点实验室 合肥 230041

通讯作者:


彭彬(1996-),男,河南驻马店人,硕士研究生,主要研究方向为深度学习车道线检测。Tel:18437903286 E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/qcgcxb/CN/10.3969/j.issn.2095–1469.2024.05.07
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