Article(id=1156908305076540161, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156908295593223005, articleNumber=null, orderNo=null, doi=10.12404/j.issn.1671-1815.2402469, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1712419200000, receivedDateStr=2024-04-07, revisedDate=1728921600000, revisedDateStr=2024-10-15, acceptedDate=null, acceptedDateStr=null, onlineDate=1753758034246, onlineDateStr=2025-07-29, pubDate=1736265600000, pubDateStr=2025-01-08, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1753758034246, onlineIssueDateStr=2025-07-29, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1753758034246, creator=13701087609, updateTime=1753758034246, updator=13701087609, issue=Issue{id=1156908295593223005, tenantId=1146029695717560320, journalId=1146123166801305609, year='2025', volume='25', issue='1', pageStart='1', pageEnd='438', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=0, createTime=1753758031985, creator=13701087609, updateTime=1765425680602, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1205845960933049001, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156908295593223005, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1205845960933049002, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156908295593223005, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=262, endPage=269, ext={EN=ArticleExt(id=1156908306506797832, articleId=1156908305076540161, tenantId=1146029695717560320, journalId=1146123166801305609, language=EN, title=High-altitude Nut Recognition Algorithm Based on Improved YOLOv5, columnId=1156262729162810294, journalTitle=Science Technology and Engineering, columnName=Papers·Automation and Computational Technology, runingTitle=null, highlight=null, articleAbstract=

In order to improve the recognition accuracy of high-altitude nuts and reduce the false detection rate of bolts and nuts, a high-altitude nut recognition model based on improved YOLOv5 was proposed. Firstly, a new attention mechanism efficient multi-scale attention(EMA) was added to the backbone network to integrate more information. Secondly, in order to enhance the network’s feature extraction capability, bidirectional feature pyramid network(BiFPN) was used to replace the PANet of the neck network. Finally, structured intersection over union(SIoU) was used to replace the original loss function complete intersection-over-union(CIoU) to accelerate the convergence of the model and improve its classification accuracy. The results show that the improved model has better performance than the original YOLOv5 model. The accuracy of the improved model increases by 0.92%. The recall increases by 0.16%. The average precision 1 (mAP_0.5:0.5) increases by 0.53%. And the average precision 2 (mAP:0.95) increases by 2.26%. An actual recognition comparison experiment between the improved model and the original YOLOv5 model was carried out. The experimental results show that the improved model has better recognition performance, which reduces the missed detection rate and the false detection rate, and improves the actual recognition rate. The improved model can well meet the recognition and image data acquisition of high-altitude nuts. And it also provide a data foundation for subsequent nut maintenance.

, correspAuthors=Wei FANG, 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=Fang-fang MENG, Xiao-zhuang TIAN, Wei FANG, Dong-ying ZHANG, Yun ZAN, Chen ZHANG, Peng ZHAN), CN=ArticleExt(id=1156908423209111760, articleId=1156908305076540161, tenantId=1146029695717560320, journalId=1146123166801305609, language=CN, title=基于改进YOLOv5的高空螺母识别算法, columnId=1156262729783567290, journalTitle=科学技术与工程, columnName=论文·自动化技术、计算机技术, runingTitle=null, highlight=null, articleAbstract=为了提升高空螺母的识别准确率,减少螺栓螺母的误检、漏检率,提出了一种基于改进YOLOv5的高空螺母识别模型。首先,在骨干网络端添加了高效多尺度注意力机制(efficient multi-scale attention,EMA),以此融合更多的信息。其次,将颈部网络的PANet更换为BiFPN(bidirectional feature pyramid network),以加强网络的特征提取能力。最后,将原损失函数CIoU(complete intersection-over-union)更换为SIoU(structured intersection over union),以加快模型的收敛速度并提高模型的分类准确率。结果表明,相比于YOLOv5原模型,改进后的模型拥有更好的性能,其中准确率提升了0.92%,召回率提升了0.16%,平均精度1(mAP_0.5)提升了0.53%,平均精度2(mAP_0.5:0.95)提升了2.26%。再用改进前后的模型进行实际识别对比实验,结果表明,改进后的模型识别效果更好,漏检、误检率下降,实际的识别率更高。改进后的模型能够很好地满足高空螺母的识别和图像数据采集,也为后续的螺母维护提供了数据基础。, correspAuthors=方薇, authorNote=null, correspAuthorsNote=
* 方薇(1977—),女,汉族,安徽合肥人,博士,副研究员。研究方向:计算机信息处理、遥感技术应用。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=Qdimq57W824PTvQN6kTdDw==, magXml=furth+PoTWV0J1qyWg0mRw==, pdfUrl=null, pdf=VqmuMYzf29wUfgH3up2Luw==, pdfFileSize=10346193, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=g4XschGCDUdB34Ej2kSFiw==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=xtr9N1/4OpyMlJgWkDLSAQ==, mapNumber=null, authorCompany=null, fund=null, authors=

孟芳芳(1984—),女,汉族,河南新乡人,博士,副教授。研究方向:智能控制、量子系统控制及应用。E-mail:

, authorsList=孟芳芳, 田孝壮, 方薇, 张冬英, 昝运, 张晨, 詹鹏)}, authors=[Author(id=1205908616784179360, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=mengffo@hfuu.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1205908616876454050, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908616784179360, language=EN, stringName=Fang-fang MENG, firstName=Fang-fang, middleName=null, lastName=MENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1205908618029887651, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908616784179360, language=CN, stringName=孟芳芳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.合肥大学先进制造工程学院, 合肥 230601, bio={"content":"

孟芳芳(1984—),女,汉族,河南新乡人,博士,副教授。研究方向:智能控制、量子系统控制及应用。E-mail:

"}, bioImg=null, bioContent=

孟芳芳(1984—),女,汉族,河南新乡人,博士,副教授。研究方向:智能控制、量子系统控制及应用。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1205908616595435673, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616603824282, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China), AuthorCompanyExt(id=1205908616612212891, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.合肥大学先进制造工程学院, 合肥 230601)])]), Author(id=1205908618113773733, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, 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=1205908618189271207, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618113773733, language=EN, stringName=Xiao-zhuang TIAN, firstName=Xiao-zhuang, middleName=null, lastName=TIAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1205908618252185768, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618113773733, language=CN, stringName=田孝壮, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.合肥大学先进制造工程学院, 合肥 230601, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1205908616595435673, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616603824282, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China), AuthorCompanyExt(id=1205908616612212891, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.合肥大学先进制造工程学院, 合肥 230601)])]), Author(id=1205908618323488938, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wfang@iim.ac.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1205908618398986412, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618323488938, language=EN, stringName=Wei FANG, firstName=Wei, middleName=null, lastName=FANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, *, address=2. Institute of Intelligent Machinery, Hefei Institute of Material Science, Chinese Academy of Sciences, Hefei 230031, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1205908618478678189, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618323488938, language=CN, stringName=方薇, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, *, address=2.中科院合肥物质科学研究院智能机械研究所, 合肥 230031, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1205908616704487580, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616717070493, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616704487580, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. Institute of Intelligent Machinery, Hefei Institute of Material Science, Chinese Academy of Sciences, Hefei 230031, China), AuthorCompanyExt(id=1205908616721264798, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616704487580, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中科院合肥物质科学研究院智能机械研究所, 合肥 230031)])]), Author(id=1205908618549981359, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, 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=1205908618621284529, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618549981359, language=EN, stringName=Dong-ying ZHANG, firstName=Dong-ying, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2. Institute of Intelligent Machinery, Hefei Institute of Material Science, Chinese Academy of Sciences, Hefei 230031, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1205908618680004786, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618549981359, language=CN, stringName=张冬英, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.中科院合肥物质科学研究院智能机械研究所, 合肥 230031, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1205908616704487580, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616717070493, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616704487580, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. Institute of Intelligent Machinery, Hefei Institute of Material Science, Chinese Academy of Sciences, Hefei 230031, China), AuthorCompanyExt(id=1205908616721264798, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616704487580, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中科院合肥物质科学研究院智能机械研究所, 合肥 230031)])]), Author(id=1205908618751307956, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, orderNo=4, 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=1205908618847776950, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618751307956, language=EN, stringName=Yun ZAN, firstName=Yun, middleName=null, lastName=ZAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1205908618906497207, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618751307956, language=CN, stringName=昝运, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.合肥大学先进制造工程学院, 合肥 230601, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1205908616595435673, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616603824282, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China), AuthorCompanyExt(id=1205908616612212891, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.合肥大学先进制造工程学院, 合肥 230601)])]), Author(id=1205908618969411769, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, orderNo=5, 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=1205908619032326331, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618969411769, language=EN, stringName=Chen ZHANG, firstName=Chen, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1205908619099435196, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908618969411769, language=CN, stringName=张晨, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.合肥大学先进制造工程学院, 合肥 230601, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1205908616595435673, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616603824282, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China), AuthorCompanyExt(id=1205908616612212891, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.合肥大学先进制造工程学院, 合肥 230601)])]), Author(id=1205908619162349758, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, orderNo=6, 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=1205908619237847232, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908619162349758, language=EN, stringName=Peng ZHAN, firstName=Peng, middleName=null, lastName=ZHAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1205908619330121921, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, authorId=1205908619162349758, language=CN, stringName=詹鹏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.合肥大学先进制造工程学院, 合肥 230601, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1205908616595435673, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616603824282, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China), AuthorCompanyExt(id=1205908616612212891, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.合肥大学先进制造工程学院, 合肥 230601)])])], keywords=[Keyword(id=1205908619476922562, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, orderNo=1, keyword=nut), Keyword(id=1205908619539837123, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, orderNo=2, keyword=object detection), Keyword(id=1205908619611140292, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, orderNo=3, keyword=YOLOv5), Keyword(id=1205908619682443461, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, orderNo=4, keyword=attention mechanism), Keyword(id=1205908619736969414, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, orderNo=1, keyword=螺母), Keyword(id=1205908619795689671, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, orderNo=2, keyword=目标检测), Keyword(id=1205908619896352968, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, orderNo=3, keyword=YOLOv5), Keyword(id=1205908619963461833, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, orderNo=4, keyword=注意力机制)], refs=[Reference(id=1205908623272767715, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=20, pageStart=90, pageEnd=94, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=黄志成, 吴永良, 梁臣, journalName=中国建设信息化, refType=null, unstructuredReference=黄志成, 吴永良, 梁臣. 基于BIM技术的格构式塔吊基础设计应用[J]. 中国建设信息化, 2023(20): 90-94., articleTitle=基于BIM技术的格构式塔吊基础设计应用, refAbstract=null), Reference(id=1205908623381819620, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=20, pageStart=90, pageEnd=94, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=Huang Zhicheng, Wu Yongliang, Liang Chen, journalName=China Construction Informatization, refType=null, unstructuredReference=Huang Zhicheng, Wu Yongliang, Liang Chen. Design and application of lattice tower crane foundation based on BIM technology[J]. China Construction Informatization, 2023(20): 90-94., articleTitle=Design and application of lattice tower crane foundation based on BIM technology, refAbstract=null), Reference(id=1205908623453122789, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=64, issue=2, pageStart=198, pageEnd=204, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=张知田, 王园园, 罗柱邦, journalName=清华大学学报(自然科学版), refType=null, unstructuredReference=张知田, 王园园, 罗柱邦, 等. 塔吊与工人空间交互下危险场景自动检测[J]. 清华大学学报(自然科学版), 2024, 64 (2): 198-204., articleTitle=塔吊与工人空间交互下危险场景自动检测, refAbstract=null), Reference(id=1205908623520231654, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=64, issue=2, pageStart=198, pageEnd=204, url=null, language=null, rfNumber=[2], rfOrder=3, authorNames=Zhang Zhitian, Wang Yuanyuan, Luo Zhubang, journalName=Journal of Tsinghua University (Natural Science Edition), refType=null, unstructuredReference=Zhang Zhitian, Wang Yuanyuan, Luo Zhubang, et al. Automatic detection of hazardous scenes under the interaction between tower cranes and worker spaces[J]. Journal of Tsinghua University (Natural Science Edition), 2024, 64 (2): 198-204., articleTitle=Automatic detection of hazardous scenes under the interaction between tower cranes and worker spaces, refAbstract=null), Reference(id=1205908623587340519, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=贾世娜, journalName=基于改进YOLOv5的小目标检测算法研究, refType=null, unstructuredReference=贾世娜. 基于改进YOLOv5的小目标检测算法研究[D]. 南昌: 南昌大学, 2022., articleTitle=null, refAbstract=null), Reference(id=1205908623688003816, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=5, authorNames=Jia Shina, journalName=Research on small object detection algorithm based on improved YOLOv5, refType=null, unstructuredReference=Jia Shina. Research on small object detection algorithm based on improved YOLOv5[D]. Nanchang: Nanchang University, 2022., articleTitle=null, refAbstract=null), Reference(id=1205908623784472809, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=420, issue=1, pageStart=135534, pageEnd=null, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=Yu F, Cai H, Zhang H, journalName=Construction and Building Materials, refType=null, unstructuredReference=Yu F, Cai H, Zhang H, et al. Automatic identification method for three-phase structure of pervious concrete based on deep learning network of Mask R-CNN[J]. Construction and Building Materials, 2024, 420(1): 135534., articleTitle=Automatic identification method for three-phase structure of pervious concrete based on deep learning network of Mask R-CNN, refAbstract=null), Reference(id=1205908623876747498, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=9, issue=1, pageStart=00449, pageEnd=null, url=null, language=null, rfNumber=[5], rfOrder=7, authorNames=Zhang Y, journalName=Applied Mathematics and Nonlinear Sciences, refType=null, unstructuredReference=Zhang Y. Research on fabric yarn detection based on improved Fast R-CNN algorithm[J]. Applied Mathematics and Nonlinear Sciences, 2024, 9(1): 00449., articleTitle=Research on fabric yarn detection based on improved Fast R-CNN algorithm, refAbstract=null), Reference(id=1205908623935467755, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=7, issue=1, pageStart=070105, pageEnd=null, url=null, language=null, rfNumber=[6], rfOrder=8, authorNames=Yang Y, journalName=Journal of Artificial Intelligence Practice, refType=null, unstructuredReference=Yang Y. Vehicle target detection algorithm based on improved Faster R-CNN for remote sensing images[J]. Journal of Artificial Intelligence Practice, 2024, 7(1): 070105., articleTitle=Vehicle target detection algorithm based on improved Faster R-CNN for remote sensing images, refAbstract=null), Reference(id=1205908624006770924, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=13, issue=3, pageStart=13031330, pageEnd=null, url=null, language=null, rfNumber=[7], rfOrder=9, authorNames=SangHyun L, Gao G, journalName=Applied Sciences, refType=null, unstructuredReference=SangHyun L, Gao G. A study on pine larva detection system using swin transformer and cascade R-CNN hybrid model[J]. Applied Sciences, 2023, 13(3): 13031330., articleTitle=A study on pine larva detection system using swin transformer and cascade R-CNN hybrid model, refAbstract=null), Reference(id=1205908624090657005, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=24, issue=9, pageStart=1007, pageEnd=1017, url=null, language=null, rfNumber=[8], rfOrder=10, authorNames=Nguyen T N, Thanh Q D B, Tran C N N, journalName=International Journal of Construction Management, refType=null, unstructuredReference=Nguyen T N, Thanh Q D B, Tran C N N, et al. Improved detection network model based on YOLOv5 for warning safety in construction sites[J]. International Journal of Construction Management, 2024, 24(9): 1007-1017., articleTitle=Improved detection network model based on YOLOv5 for warning safety in construction sites, refAbstract=null), Reference(id=1205908624157765870, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=13, issue=29, pageStart=145, pageEnd=148, url=null, language=null, rfNumber=[9], rfOrder=11, authorNames=王景鑫, 潘欣, journalName=科技创新与应用, refType=null, unstructuredReference=王景鑫, 潘欣. 一种基于Labelimg的辅助标注方法[J]. 科技创新与应用, 2023, 13(29): 145-148., articleTitle=一种基于Labelimg的辅助标注方法, refAbstract=null), Reference(id=1205908624233263343, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=13, issue=29, pageStart=145, pageEnd=148, url=null, language=null, rfNumber=[9], rfOrder=12, authorNames=Wang Jingxin, Pan Xin, journalName=Technological Innovation and Application, refType=null, unstructuredReference=Wang Jingxin, Pan Xin. A labelimg based auxiliary annotation method[J]. Technological Innovation and Application, 2023, 13(29): 145-148., articleTitle=A labelimg based auxiliary annotation method, refAbstract=null), Reference(id=1205908624300372208, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=1905, pageEnd=null, url=null, language=null, rfNumber=[10], rfOrder=13, authorNames=Zou Z X, Shi Z W, Guo Y H, journalName=ArXiv, refType=null, unstructuredReference=Zou Z X, Shi Z W, Guo Y H, et al. Object detection in 20 years: a survey[J]. ArXiv, 2019: 1905.05055., articleTitle=Object detection in 20 years: a survey, refAbstract=null), Reference(id=1205908624375869681, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=59, issue=18, pageStart=436, pageEnd=446, url=null, language=null, rfNumber=[11], rfOrder=14, authorNames=马永康, 刘华, 凌成星, journalName=激光与光电子学进展, refType=null, unstructuredReference=马永康, 刘华, 凌成星, 等. 基于改进YOLOv5的红树林单木目标检测研究[J]. 激光与光电子学进展, 2022, 59(18): 436-446., articleTitle=基于改进YOLOv5的红树林单木目标检测研究, refAbstract=null), Reference(id=1205908624430395634, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=59, issue=18, pageStart=436, pageEnd=446, url=null, language=null, rfNumber=[11], rfOrder=15, authorNames=Ma Yongkang, Liu Hua, Ling Chengxing, journalName=Progress in Laser and Optoelectronics, refType=null, unstructuredReference=Ma Yongkang, Liu Hua, Ling Chengxing, et al. Research on mangrove single tree target detection based on improved YOLOv5[J]. Progress in Laser and Optoelectronics, 2022, 59 (18): 436-446., articleTitle=Research on mangrove single tree target detection based on improved YOLOv5, refAbstract=null), Reference(id=1205908624493310195, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=13708, pageEnd=13717, url=null, language=null, rfNumber=[12], rfOrder=16, authorNames=Hou Q, Zhou D, Feng J, Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR, journalName=null, refType=null, unstructuredReference=Hou Q, Zhou D, Feng J. Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition ( CVPR).Coordinate attention for efficient mobile network design[C]// Piscataway: IEEE, 2021: 13708-13717., articleTitle=null, refAbstract=null), Reference(id=1205908624644305140, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=[13], rfOrder=17, authorNames=Ouyang D, He S, Zhang G, journalName=Proceedings of the ICASSP2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), refType=null, unstructuredReference=Ouyang D, He S, Zhang G, et al. Efficient multi-scale attention module with cross-spatial learning[C]// Proceedings of the ICASSP2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Rhodes: IEEE, 2023: 1-5., articleTitle=Efficient multi-scale attention module with cross-spatial learning, refAbstract=null), Reference(id=1205908624719802613, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=39, issue=23, pageStart=174, pageEnd=184, url=null, language=null, rfNumber=[14], rfOrder=18, authorNames=李韬, 任玲, 胡斌, journalName=农业工程学报, refType=null, unstructuredReference=李韬, 任玲, 胡斌, 等. 改进YOLOv5s和迁移学习对番茄穴盘苗的分级检测[J]. 农业工程学报, 2023, 39(23): 174-184., articleTitle=改进YOLOv5s和迁移学习对番茄穴盘苗的分级检测, refAbstract=null), Reference(id=1205908624807882998, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=39, issue=23, pageStart=174, pageEnd=184, url=null, language=null, rfNumber=[14], rfOrder=19, authorNames=Li Tao, Ren Ling, Hu Bin, journalName=Journal of Agricultural Engineering, refType=null, unstructuredReference=Li Tao, Ren Ling, Hu Bin, et al. Improving YOLOv5s and transfer learning for grading detection of tomato hole seedlings[J]. Journal of Agricultural Engineering, 2023, 39(23): 174-184., articleTitle=Improving YOLOv5s and transfer learning for grading detection of tomato hole seedlings, refAbstract=null), Reference(id=1205908624866603255, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=31, issue=1, pageStart=54, pageEnd=56, url=null, language=null, rfNumber=[15], rfOrder=20, authorNames=陈雷平, 段意强, 瞿宏愿, journalName=技术与市场, refType=null, unstructuredReference=陈雷平, 段意强, 瞿宏愿. 基于改进YOLOv5的输电线路缺陷检测方法[J]. 技术与市场, 2024, 31(1): 54-56, 62., articleTitle=基于改进YOLOv5的输电线路缺陷检测方法, refAbstract=null), Reference(id=1205908624921129208, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=31, issue=1, pageStart=54, pageEnd=56, url=null, language=null, rfNumber=[15], rfOrder=21, authorNames=Chen Leiping, Duan Yiqiang, Qu Hongyuan, journalName=Technology and Market, refType=null, unstructuredReference=Chen Leiping, Duan Yiqiang, Qu Hongyuan. A defect detection method for transmission lines based on improved YOLOv5[J]. Technology and Market, 2024, 31(1): 54-56, 62., articleTitle=A defect detection method for transmission lines based on improved YOLOv5, refAbstract=null), Reference(id=1205908625017598201, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=10778, pageEnd=10787, url=null, language=null, rfNumber=[16], rfOrder=22, authorNames=Tan M, Pang R, Le Q. V., journalName=null, refType=null, unstructuredReference=Tan M, Pang R, Le Q. V. Efficientdet:scalable and efficient object detection[C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition ( CVPR). Piscataway: IEEE, 2019: 10778-10787., articleTitle=Efficientdet:scalable and efficient object detection[C], refAbstract=null), Reference(id=1205908625084707066, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=2205, pageEnd=null, url=null, language=null, rfNumber=[17], rfOrder=23, authorNames=Gevorgyan Z, journalName=ArXiv, refType=null, unstructuredReference=Gevorgyan Z. SIoU loss: more powerful learning for bounding box regression[J]. ArXiv, 2022: 2205.12740., articleTitle=SIoU loss: more powerful learning for bounding box regression, refAbstract=null), Reference(id=1205908625160204539, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=47, issue=3, pageStart=179, pageEnd=186, url=null, language=null, rfNumber=[18], rfOrder=24, authorNames=董恒祥, 潘江如, 董芙楠, journalName=现代电子技术, refType=null, unstructuredReference=董恒祥, 潘江如, 董芙楠, 等. 基于YOLOv5s模型的边界框回归损失函数研究[J]. 现代电子技术, 2024, 47 (3): 179-186., articleTitle=基于YOLOv5s模型的边界框回归损失函数研究, refAbstract=null), Reference(id=1205908625231507708, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2024, volume=47, issue=3, pageStart=179, pageEnd=186, url=null, language=null, rfNumber=[18], rfOrder=25, authorNames=Dong Hengxiang, Pan Jiangru, Dong Funan, journalName=Modern Electronic Technology, refType=null, unstructuredReference=Dong Hengxiang, Pan Jiangru, Dong Funan, et al. Research on bounding box regression loss function based on YOLOv5s model[J]. Modern Electronic Technology, 2024, 47(3): 179-186., articleTitle=Research on bounding box regression loss function based on YOLOv5s model, refAbstract=null), Reference(id=1205908625294422269, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=39, issue=18, pageStart=192, pageEnd=200, url=null, language=null, rfNumber=[19], rfOrder=26, authorNames=孙俊, 吴兆祺, 贾忆琳, journalName=农业工程学报, refType=null, unstructuredReference=孙俊, 吴兆祺, 贾忆琳, 等. 基于改进YOLOv5s的果园环境葡萄检测[J]. 农业工程学报, 2023, 39(18): 192-200., articleTitle=基于改进YOLOv5s的果园环境葡萄检测, refAbstract=null), Reference(id=1205908625390891262, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2023, volume=39, issue=18, pageStart=192, pageEnd=200, url=null, language=null, rfNumber=[19], rfOrder=27, authorNames=Sun Jun, Wu Zhaoqi, Jia Yilin, journalName=Journal of Agricultural Engineering, refType=null, unstructuredReference=Sun Jun, Wu Zhaoqi, Jia Yilin, et al. Grape detection in orchard environment based on improved YOLOv5s[J]. Journal of Agricultural Engineering, 2023, 39(18): 192-200., articleTitle=Grape detection in orchard environment based on improved YOLOv5s, refAbstract=null), Reference(id=1205908625470583039, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=53, issue=S1, pageStart=232, pageEnd=240, url=null, language=null, rfNumber=[20], rfOrder=28, authorNames=张志远, 罗铭毅, 郭树欣, journalName=农业机械学报, refType=null, unstructuredReference=张志远, 罗铭毅, 郭树欣, 等. 基于改进YOLOv5的自然环境下樱桃果实识别方法[J]. 农业机械学报, 2022, 53(S1): 232-240., articleTitle=基于改进YOLOv5的自然环境下樱桃果实识别方法, refAbstract=null), Reference(id=1205908625546080512, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=53, issue=S1, pageStart=232, pageEnd=240, url=null, language=null, rfNumber=[20], rfOrder=29, authorNames=Zhang Zhiyuan, Luo Mingyi, Guo Shuxin, journalName=Journal of Agricultural Machinery, refType=null, unstructuredReference=Zhang Zhiyuan, Luo Mingyi, Guo Shuxin, et al. Cherry fruit recognition method in natural environment based on improved YOLOv5[J]. Journal of Agricultural Machinery, 2022, 53(S1): 232-240., articleTitle=Cherry fruit recognition method in natural environment based on improved YOLOv5, refAbstract=null), Reference(id=1205908625617383681, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=22, issue=30, pageStart=13406, pageEnd=13416, url=null, language=null, rfNumber=[21], rfOrder=30, authorNames=赵桂平, 邓飞, 王昀, journalName=科学技术与工程, refType=null, unstructuredReference=赵桂平, 邓飞, 王昀, 等. 改进的YOLOv5-ResNet相似目标检测方法[J]. 科学技术与工程, 2022, 22(30): 13406-13416., articleTitle=改进的YOLOv5-ResNet相似目标检测方法, refAbstract=null), Reference(id=1205908625755795714, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, doi=null, pmid=null, pmcid=null, year=2022, volume=22, issue=30, pageStart=13406, pageEnd=13416, url=null, language=null, rfNumber=[21], rfOrder=31, authorNames=Zhao Guiping, Deng Fei, Wang Yun, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Zhao Guiping, Deng Fei, Wang Yun, et al. Improved YOLOv5 ResNet similar object detection method[J]. Science Technology and Engineering, 2022, 22(30): 13406-13416., articleTitle=Improved YOLOv5 ResNet similar object detection method, refAbstract=null)], funds=[Fund(id=1205908623121772770, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, awardId=61973290, language=CN, fundingSource=国家自然科学基金(61973290), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1205908616595435673, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616603824282, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China), AuthorCompanyExt(id=1205908616612212891, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616595435673, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.合肥大学先进制造工程学院, 合肥 230601)]), AuthorCompany(id=1205908616704487580, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, xref=null, ext=[AuthorCompanyExt(id=1205908616717070493, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616704487580, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2. Institute of Intelligent Machinery, Hefei Institute of Material Science, Chinese Academy of Sciences, Hefei 230031, China), AuthorCompanyExt(id=1205908616721264798, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, companyId=1205908616704487580, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.中科院合肥物质科学研究院智能机械研究所, 合肥 230031)])], figs=[ArticleFig(id=1205908620127039690, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.1, caption=Dirty nut image, figureFileSmall=a9VOtoN/zco7nVMGiV5FQg==, figureFileBig=7PzDp+05ciIlPbdWuTr7kA==, tableContent=null), ArticleFig(id=1205908620194148555, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图1, caption=脏污螺母图像, figureFileSmall=a9VOtoN/zco7nVMGiV5FQg==, figureFileBig=7PzDp+05ciIlPbdWuTr7kA==, tableContent=null), ArticleFig(id=1205908620273840332, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.2, caption=Schematic diagram of nut and bolt labeling, figureFileSmall=7WZXWFvnbGD0x8ekEiDetA==, figureFileBig=WAHzpv0QpIMotCwzMMNN3w==, tableContent=null), ArticleFig(id=1205908620349337805, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图2, caption=螺母螺栓标注示意图, figureFileSmall=7WZXWFvnbGD0x8ekEiDetA==, figureFileBig=WAHzpv0QpIMotCwzMMNN3w==, tableContent=null), ArticleFig(id=1205908620416446670, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.3, caption=Width and height distribution diagram of nut bolt boundary, figureFileSmall=rIvstpMcXzvfx8jYrVzhfg==, figureFileBig=+Gb1/02qCYYiXsbHuICoDQ==, tableContent=null), ArticleFig(id=1205908620496138447, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图3, caption=螺母螺栓边界宽高分布图, figureFileSmall=rIvstpMcXzvfx8jYrVzhfg==, figureFileBig=+Gb1/02qCYYiXsbHuICoDQ==, tableContent=null), ArticleFig(id=1205908620571635920, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.4, caption=YOLOv5 network mode, figureFileSmall=p729WJOOnLRHS4UX1XL4bw==, figureFileBig=ow0LaXgFLeMGKUlwOpFoHQ==, tableContent=null), ArticleFig(id=1205908620647133393, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图4, caption=YOLOv5网络模型, figureFileSmall=p729WJOOnLRHS4UX1XL4bw==, figureFileBig=ow0LaXgFLeMGKUlwOpFoHQ==, tableContent=null), ArticleFig(id=1205908620718436562, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.5, caption=EMA network architecture diagram, figureFileSmall=huT4kcSr+2KUV3Uw0l3VGA==, figureFileBig=cjmLKh6QlQMULItgwhEEhg==, tableContent=null), ArticleFig(id=1205908620785545427, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图5, caption=EMA网络结构图, figureFileSmall=huT4kcSr+2KUV3Uw0l3VGA==, figureFileBig=cjmLKh6QlQMULItgwhEEhg==, tableContent=null), ArticleFig(id=1205908620873625812, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.6, caption=Comparison diagram of Neck network structure, figureFileSmall=zcDl+KKG0lguLA17bll/0w==, figureFileBig=QQv4iP9BkIskSi3gKqOe3w==, tableContent=null), ArticleFig(id=1205908620957511893, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图6, caption=Neck网络结构对比图, figureFileSmall=zcDl+KKG0lguLA17bll/0w==, figureFileBig=QQv4iP9BkIskSi3gKqOe3w==, tableContent=null), ArticleFig(id=1205908621024620758, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.7, caption=Angle loss diagram, figureFileSmall=lwq4KRIOW++A3qKiHJMvfA==, figureFileBig=79XMaP88Ul0BWoX6gwq+eQ==, tableContent=null), ArticleFig(id=1205908621091729623, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图7, caption=角度损失图

B为真实框;BGT为预测框;cwch分别为真实框和预测框最小外接矩形的宽和高;ασcw的夹角;βσch的夹角

, figureFileSmall=lwq4KRIOW++A3qKiHJMvfA==, figureFileBig=79XMaP88Ul0BWoX6gwq+eQ==, tableContent=null), ArticleFig(id=1205908621175615704, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.8, caption=Shape loss diagram, figureFileSmall=FsY/jE8WmNUdL68vHE+zNA==, figureFileBig=gdCa4daAb7xrKnZ2ugkPAg==, tableContent=null), ArticleFig(id=1205908621263696089, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图8, caption=形状损失图, figureFileSmall=FsY/jE8WmNUdL68vHE+zNA==, figureFileBig=gdCa4daAb7xrKnZ2ugkPAg==, tableContent=null), ArticleFig(id=1205908621334999258, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.9, caption=Parameter variation curves of different YOLO improved algorithms, figureFileSmall=g0rqxvTcXgOlFYklKOW8pA==, figureFileBig=Rv9KcUajHtkT25iIkai6og==, tableContent=null), ArticleFig(id=1205908622501015771, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图9, caption=不同YOLO改进算法的参数变化曲线图, figureFileSmall=g0rqxvTcXgOlFYklKOW8pA==, figureFileBig=Rv9KcUajHtkT25iIkai6og==, tableContent=null), ArticleFig(id=1205908622618456284, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Fig.10, caption=Comparison diagram of improved model nut detection, figureFileSmall=wqbCe7QP+F+tpF8+1NhDtw==, figureFileBig=m/uRIXqVyd4Kl4cMYBIhog==, tableContent=null), ArticleFig(id=1205908622685565149, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=图10, caption=改进模型螺母检测对比图, figureFileSmall=wqbCe7QP+F+tpF8+1NhDtw==, figureFileBig=m/uRIXqVyd4Kl4cMYBIhog==, tableContent=null), ArticleFig(id=1205908622748479710, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Table 1, caption=

Experimental platform parameters

, figureFileSmall=null, figureFileBig=null, tableContent=
环境 参数
操作系统 Windows 11
CPU I9-11900K 3.5 GHz
GPU RTX3060Ti
显存 8 G
内存 128 G
框架 Pytorch 2.0.1+CUDA 12.3
编程语言 Python 3.10.12
), ArticleFig(id=1205908622828171487, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=表1, caption=

实验平台参数

, figureFileSmall=null, figureFileBig=null, tableContent=
环境 参数
操作系统 Windows 11
CPU I9-11900K 3.5 GHz
GPU RTX3060Ti
显存 8 G
内存 128 G
框架 Pytorch 2.0.1+CUDA 12.3
编程语言 Python 3.10.12
), ArticleFig(id=1205908622899474656, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=EN, label=Table 2, caption=

Results of ablation experiment

, figureFileSmall=null, figureFileBig=null, tableContent=
实验
编号
EMA BiFPN SIoU 准确
率/%
召回
率/%
平均精
度1/%
平均精
度2/%
1 × × × 98.05 97.74 98.50 78.12
2 × × 98.85 97.48 98.86 79.62
3 × × 98.60 97.61 98.89 79.16
4 × × 97.93 97.44 98.76 79.48
5 × 98.43 97.81 98.90 79.53
6 × 98.61 97.44 98.96 80.24
7 × 98.50 97.77 98.85 79.98
8 98.97 97.90 99.03 80.38
), ArticleFig(id=1205908622979166433, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156908305076540161, language=CN, label=表2, caption=

消融实验结果

, figureFileSmall=null, figureFileBig=null, tableContent=
实验
编号
EMA BiFPN SIoU 准确
率/%
召回
率/%
平均精
度1/%
平均精
度2/%
1 × × × 98.05 97.74 98.50 78.12
2 × × 98.85 97.48 98.86 79.62
3 × × 98.60 97.61 98.89 79.16
4 × × 97.93 97.44 98.76 79.48
5 × 98.43 97.81 98.90 79.53
6 × 98.61 97.44 98.96 80.24
7 × 98.50 97.77 98.85 79.98
8 98.97 97.90 99.03 80.38
)], attaches=null, journal=Journal(id=1146119176004939786, delFlag=0, nameCn=科学技术与工程, nameEn=Science Technology and Engineering, nameHistory1=null, nameHistory2=null, issn=1671-1815, eissn=, cn=11-4688/T, coden=null, periodic=4, language=CN, oaType=是, 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=UKU/O7GSka5polgCTkbIIw==, journalPrice=null, startedYear=null, abbrevIsoEn=Sci Technol Eng, journalRemark=null, publicationField=null, createdTime=null, updatedTime=1754445529766, createdBy=null, updatedBy=13701087609, firstLetterCn=S, firstLetterEn=S, subjectCode=Natural Sciences, subjectName=自然科学, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=UKU/O7GSka5polgCTkbIIw==, picEn=5hwlULoNwcbj3xUmVi9MAQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1159791870395564357, language=CN, name=科学技术与工程, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.stae.com.cn/jsygc/home, createdTime=1754445529793, updatedTime=1754445529793, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=http://www.stae.com.cn/jsygc/site/menus/20090429150146001, submissionAuthorUrl=http://www.stae.com.cn/jsygc/author/login, submissionEditorUrl=http://www.stae.com.cn/jsygc/editor/login, submissionReviewUrl=http://www.stae.com.cn/jsygc/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1159791870441701702, language=EN, name=Science Technology and Engineering, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=null, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=http://www.stae.com.cn/jsygc/home, createdTime=1754445529804, updatedTime=1754445529804, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://www.stae.com.cn/jsygc/author/login, submissionEditorUrl=http://www.stae.com.cn/jsygc/editor/login, submissionReviewUrl=http://www.stae.com.cn/jsygc/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1146123166801305609, websiteList=[Website(id=1148243202391400884, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146123166801305609, 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/kxjsygc/CN, language=CN, createTime=1751692112777, createBy=18614031015, updateTime=1753520965431, updateBy=18614031015, name=科学技术与工程-中文站点, tplId=1146099689490845704, title=科学技术与工程, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1148622798802673703, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=articleTextType, value=kx, createTime=1751782615614, updateTime=1751782615614, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798781702180, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=banner, value=null, createTime=1751782615609, updateTime=1751782615609, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798769119267, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=j86gbwi+p0Idkyl5SzIlmQ==, createTime=1751782615606, updateTime=1751782615606, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798794285094, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1751782615612, updateTime=1751782615612, creator=18614031015, updator=18614031015), WebsiteProps(id=1148622798790090789, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1148243202391400884, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1751782615611, updateTime=1751782615611, creator=18614031015, updator=18614031015)]), Website(id=1155914124811976731, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1146123166801305609, 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/kxjsygc/EN, language=EN, createTime=1753521003206, createBy=18614031015, updateTime=1753521003206, updateBy=18614031015, name=科学技术与工程-英文站点, tplId=1146101810881728533, title=Science Technology and Engineering, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1155914371227308235, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=articleTextType, value=kx, createTime=1753521061952, updateTime=1753521061952, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371210531016, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=banner, value=null, createTime=1753521061947, updateTime=1753521061947, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371202142407, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=logo, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic?fileId=j86gbwi+p0Idkyl5SzIlmQ==, createTime=1753521061945, updateTime=1753521061945, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371223113930, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/kjdb/CN/file/pic, createTime=1753521061950, updateTime=1753521061950, creator=18614031015, updator=18614031015), WebsiteProps(id=1155914371218919625, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1155914124811976731, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1753521061949, updateTime=1753521061949, creator=18614031015, updator=18614031015)])], journalTitle=科学技术与工程, weixinUrl=null, journalUrl=null, iacademicId=null, status=0, seqNo=null, journalTitleEn=Science Technology and Engineering, journalPhotoCn=UKU/O7GSka5polgCTkbIIw==, journalPhotoEn=5hwlULoNwcbj3xUmVi9MAQ==, journalFirstLetter=S, 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=null, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2402469, detailUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/10.12404/j.issn.1671-1815.2402469, pdfUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/PDF/10.12404/j.issn.1671-1815.2402469, pdfUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/PDF/10.12404/j.issn.1671-1815.2402469, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于改进YOLOv5的高空螺母识别算法
收藏切换
PDF下载
孟芳芳 1 , 田孝壮 1 , 方薇 2, * , 张冬英 2 , 昝运 1 , 张晨 1 , 詹鹏 1
科学技术与工程 | 论文·自动化技术、计算机技术 2025,25(1): 262-269
收起
收藏切换
科学技术与工程 | 论文·自动化技术、计算机技术 2025, 25(1): 262-269
基于改进YOLOv5的高空螺母识别算法
全屏
孟芳芳1 , 田孝壮1, 方薇2, * , 张冬英2, 昝运1, 张晨1, 詹鹏1
作者信息
  • 1.合肥大学先进制造工程学院, 合肥 230601
  • 2.中科院合肥物质科学研究院智能机械研究所, 合肥 230031
  • 孟芳芳(1984—),女,汉族,河南新乡人,博士,副教授。研究方向:智能控制、量子系统控制及应用。E-mail:

通讯作者:

* 方薇(1977—),女,汉族,安徽合肥人,博士,副研究员。研究方向:计算机信息处理、遥感技术应用。E-mail:
High-altitude Nut Recognition Algorithm Based on Improved YOLOv5
Fang-fang MENG1 , Xiao-zhuang TIAN1, Wei FANG2, * , Dong-ying ZHANG2, Yun ZAN1, Chen ZHANG1, Peng ZHAN1
Affiliations
  • 1. School of Advanced Manufacturing Engineering, Hefei University, Hefei 230601, China
  • 2. Institute of Intelligent Machinery, Hefei Institute of Material Science, Chinese Academy of Sciences, Hefei 230031, China
出版时间: 2025-01-08 doi: 10.12404/j.issn.1671-1815.2402469
文章导航
收藏切换
为了提升高空螺母的识别准确率,减少螺栓螺母的误检、漏检率,提出了一种基于改进YOLOv5的高空螺母识别模型。首先,在骨干网络端添加了高效多尺度注意力机制(efficient multi-scale attention,EMA),以此融合更多的信息。其次,将颈部网络的PANet更换为BiFPN(bidirectional feature pyramid network),以加强网络的特征提取能力。最后,将原损失函数CIoU(complete intersection-over-union)更换为SIoU(structured intersection over union),以加快模型的收敛速度并提高模型的分类准确率。结果表明,相比于YOLOv5原模型,改进后的模型拥有更好的性能,其中准确率提升了0.92%,召回率提升了0.16%,平均精度1(mAP_0.5)提升了0.53%,平均精度2(mAP_0.5:0.95)提升了2.26%。再用改进前后的模型进行实际识别对比实验,结果表明,改进后的模型识别效果更好,漏检、误检率下降,实际的识别率更高。改进后的模型能够很好地满足高空螺母的识别和图像数据采集,也为后续的螺母维护提供了数据基础。
螺母  /  目标检测  /  YOLOv5  /  注意力机制

In order to improve the recognition accuracy of high-altitude nuts and reduce the false detection rate of bolts and nuts, a high-altitude nut recognition model based on improved YOLOv5 was proposed. Firstly, a new attention mechanism efficient multi-scale attention(EMA) was added to the backbone network to integrate more information. Secondly, in order to enhance the network’s feature extraction capability, bidirectional feature pyramid network(BiFPN) was used to replace the PANet of the neck network. Finally, structured intersection over union(SIoU) was used to replace the original loss function complete intersection-over-union(CIoU) to accelerate the convergence of the model and improve its classification accuracy. The results show that the improved model has better performance than the original YOLOv5 model. The accuracy of the improved model increases by 0.92%. The recall increases by 0.16%. The average precision 1 (mAP_0.5:0.5) increases by 0.53%. And the average precision 2 (mAP:0.95) increases by 2.26%. An actual recognition comparison experiment between the improved model and the original YOLOv5 model was carried out. The experimental results show that the improved model has better recognition performance, which reduces the missed detection rate and the false detection rate, and improves the actual recognition rate. The improved model can well meet the recognition and image data acquisition of high-altitude nuts. And it also provide a data foundation for subsequent nut maintenance.

nut  /  object detection  /  YOLOv5  /  attention mechanism
孟芳芳, 田孝壮, 方薇, 张冬英, 昝运, 张晨, 詹鹏. 基于改进YOLOv5的高空螺母识别算法. 科学技术与工程, 2025 , 25 (1) : 262 -269 . DOI: 10.12404/j.issn.1671-1815.2402469
Fang-fang MENG, Xiao-zhuang TIAN, Wei FANG, Dong-ying ZHANG, Yun ZAN, Chen ZHANG, Peng ZHAN. High-altitude Nut Recognition Algorithm Based on Improved YOLOv5[J]. Science Technology and Engineering, 2025 , 25 (1) : 262 -269 . DOI: 10.12404/j.issn.1671-1815.2402469
高空作业通常指人在一定位置为基准的高处进行的作业,常常伴随着危险,高空螺母的检测也是高空作业的一种。基于目标识别模型进行改进并应用在高空螺母的识别与采集中,如输电工程中的角钢塔和建筑工地中的塔吊,角钢塔和塔吊上的螺母在使用过程中可能会发生损坏或者松动,需要定期检测,该识别模型可对螺母进行识别和采集,为其后续的缺陷检测和松动检测提供前置数据。从而替代人工检测,减少人工高空作业,降低安全事故的发生,并为全自动的高空螺母识别和检测提供技术支持。
塔吊一般超过十节以上,螺栓数量众多,还分布在每一层,需要耗费大量的人力才能完成对螺栓的检测,还可能存在漏检的情况,给后期塔吊工作埋下安全隐患[1-2]
传统检修螺母的方法是由工人进行定期检修,通过人工的方式进行塔吊螺母的检测和修理效率低并且伴随着危险。随着机器视觉的发展,深度学习模型对目标的特征提取能力逐渐提高,基于深度学习的目标识别方法提高了识别的精度。但对于螺母被遮挡的情况,其识别率较低,并容易受光线等环境因素的影响,从而导致误检、漏检的情况发生,致使其识别能力不能满足实际的需求。
基于深度学习的目标检测算法可大致分为两类:一是基于候选区域的两阶段算法,另一是基于回归的一阶段算法[3]。典型的两阶段算法包括区域卷积神经网络(region-based convolutional neural network,R-CNN)[4]、Fast R-CNN[5]、Faster R-CNN[6]和Cascade R-CNN[7]。两阶段算法引入了区域候选网络(region proposal network,RPN),先进行一步预选,然后通过预选结果获取最终检测结果,提高了精确性。然而,这种算法较为复杂,在硬件较差的轻型设备中运行较慢,检测速度不理想。为了解决这个问题,一阶段检测算法应运而生,它具有检测速度快、效率高、对硬件要求低等优点。YOLO(you only look once)[8]系列模型具有良好的全局感受野,通过使用多语义融合检测机制、网格划分和锚框匹配,能够对感受野受限的问题进行有效的改善,可以实现对目标的高效率检测,其性能优秀被应用在各个领域中,YOLOv5目标检测模型就是一阶段检测算法的其中之一。为了将该方法更好地应用于高空螺母检测中,将对YOLOv5模型进行改进以实现对高空螺母的精准检测。
现以塔吊上的高空螺母作为研究对象,以YOLOv5作为基础的目标检测模型,为了提高螺母在遮挡情况下的识别率和克服天气等环境因素造成的影响和误差,对该模型进行必要的改进,在Backbone端添加跨空间学习的高效多尺度注意力机制(efficient multi-scale attention,EMA),融合多尺度的信息,减少干扰;在Neck端增加了跨尺度加权特征融合-BiFPN(bidirectional feature pyramid network),以增强不同尺度的特征融合能力,更换损失函数为SIoU(structured intersection over union),提高模型收敛速度。改进后的模型将有助于提高被遮挡的高空螺母的识别率和识别精度,降低漏检和误检的发生。
螺母图像采集主要在安徽合肥部分工地进行,采用大疆无人机Mavic3E拍摄采集螺母数据,无人机系统携带检测设备进行作业,实现自主飞行,取代人工携带设备行进、攀爬等过程,能降低作业难度和危险性。大疆Mavic 3E的相机为2 000万像素,4/3CMOS(complementary metal oxide semiconductor),等效焦距24 mm,像元3.3 μm,能够采集清晰的螺母图像。在不同天气、光线下均进行了采集,还采集了不同状况的螺母,如干净清晰的螺母和附着水泥、锈迹、油漆的脏污螺母,如图1所示。
同时采集的时候对同一螺母进行了不同角度的采集,扩充了数据集,包含了更多类型的数据,以保证训练的样本多样和识别的准确性。采集的图片格式为JPG,图片分辨率为5 280像素×3 956像素。采集到的塔吊图像总共1 032张,每张图像约包含6个螺母和螺栓。
对于采集到的螺母图像进行浏览筛选,去除因为对焦点错误导致螺母螺栓模糊的图像,最终得到927张塔吊图像,每张图像约包含6个螺母和螺栓,共计5 562个螺母螺栓。使用LabelImg软件对927张图像里的螺母螺栓进行标注,得到包含了螺母螺栓在图像中的中心坐标和宽高信息的xml类型文件,通过python编写代码将xml数据集格式转换为YOLO可识别的txt数据集格式,并按照8∶2划分训练集和验证集。txt文件的第一列为目标类别,后面4个数字为[x-center, y-center, w, h],即[x-中心,y-中心,宽度,高度]该数字均小于1的[9],因为对应的是整张图片的比例,所以即使图像被拉伸放缩,这种格式的标签也可以找到相应的目标,实际标注如图2所示。
对螺母螺栓标注边界的宽高进行了统计,统计结果如图3所示。可以看出,大部分的螺母螺栓的标签框纵横比接近1∶1,因为双螺母叠加后的宽和高接近,且形态不会发生变化,螺栓的在螺母上方露出度也基本相同。标注框基本保持类似的比例降低了检测模型在位置预测过拟合的情况。从图3还可以看出螺母螺栓所占的整张图片的面积比例很小,大多在2.4%以内。
YOLOv5模型网络包含输入端(Input)、骨干网络(Backbone)、特征融合网络(Neck)、预测网络(Head)[10],YOLOv5网络结构图如图4所示。
在骨干网络中YOLOv5 6.0版本用一个卷积层(6×6)替代了原有的Focus模块,替换过后的计算量不变,但使用卷积层更加高效。主干网络中还包含有CBS(conv-batchnorm2d-silu)模块、C3(由3个cbs模块和1个bottleneck模块组成)模块、SPPF(spatial pyramid pooling with features)模块。Backbone网络后接Neck网络,其为特征融合网络,作用是从Backbone中获取相对于较浅的特征,再与深层的语义特征Concat到一起。它包含了特征金字塔FPN+PAN,拥有双向融合[11],既可以捕获强语义特征,又可以传达强定位特征。Head网络为Detect模块,Detect模块的网络结构很简单,仅由3个1×1卷积构成,对应3个检测特征层,用来进行目标检测和输出检测结果。
在各种目标识别模型中添加的注意力机制,能够帮助模型识别到更多的特征,但这可能会带来一定的副作用。比如CA(coordinate attention)注意力机制[12],虽对精度有一定的提升,但是由于需要对整个特征图进行注意力权重的计算,因此会额外消耗更多的资源,并且无法捕捉通道之间长距离的依赖关系。现引入了新型的高效多尺度EMA[13]注意力模块,更加关注感兴趣区域,提高了主要特征的权重,并降低了次要特征的权重,有助于减少天气等环境因素对识别造成的影响,从而减少误检、漏检的发生,EMA网络结构图如图5所示。
EMA注意力机制具有1×1分支和3×3分支两条路线,为了聚合多尺度空间结构信息,将两条路线并行放置,以实现快速响应。首先EMA将输入特征图X∈RC×H×W分成沿通道维度方向的G个子特征,以用来学习不同的语义特征,可由X=[X0,X1,…,XG-1]表示[14],EMA通过1×1分支和3×3分支来提取特征图的注意力权重描述符。
再将两组的通道注意图通过一个简单的计算,使它们聚合在一起,从而在1×1分支的两条空间方向路线中实现不同的跨通道交互特征。在3×3分支上通过一个3×3卷积来捕捉多尺度特征。在该通道方向上建立跨通道信息交互模型,用来捕获所有通道之间的依赖关系并能有效地减少计算量,从而扩大特征空间。
EMA引入了两个张量,分别为1×1分支的输出和3×3分支的输出。在1×1分支中利用二维全局平均池化将全局空间信息编码到1×1分支的输出中,并将最小分支的输出直接转换为相应维度的形状,随后将二维全局平均池化的输出通过二维非线性函数Softmax将上面的线性变换进行拟合。通过将上述并行处理的输出与矩阵点积运算相乘,得到了第一张空间注意力图,并同时收集了不同尺度的空间信息。3×3分支与1×1分支类似,经过相同步骤得出第二个空间注意力图,该图精确地保留了整个空间位置信息。最后用Sigmoid函数进行计算,EMA的最终输出与X大小相同,其不仅可以对通道间信息进行编码以调整不同通道的重要性,还可以将精确的空间结构信息保留到通道中,从而可以高效地堆叠到YOLOv5架构中。
YOLOv5在Neck网络使用的是FPN+PANet(feature pyramid network + path aggregation network),FPN结构是自下而上捕获语义特征,使用图像上采样向特征图中进行插值,使特征图的尺寸变大,以便于融合来自Backbone网络的特征图,进行特征的向上融合。但是它仅仅增强了图像中的语义信息,却忽略了定位信息。而PANet对定位信息的传递进行了补充,它在FPN的结构上增加了一个自下而上的特征金字塔,通过这个特征金字塔将下层的定位信息向上层进行传递,这种具有双向融合的结构为PANet,它既结合了语义信息又包含了定位信息[15]。不过这种双向融合的结构较为简单,融合的特征不够多。故引入了具有跨尺度连接和加权特征融合的BiFPN[16]。由于节点是单输入的且没有特征融合,则它的贡献对特征融合网络来说就非常小,BiFPN去掉了单输入的节点,这样便可在不损失很多精度的前提下简化特征网络。它还在同一水平的输入节点和输出节点添加了一条额外的路径,以此融合了更多的特征,并且没有增加太多的计算量。BiFPN融合了更多的语义特征信息,这有助于提升遮挡目标的识别率,减少漏检的情况发生。3种Neck网络结构如图6所示。
对于螺母螺栓识别来说,BiFPN采用跨尺度连接方式对不同的检测特征通过跨尺度权重进行相对应的表达,从而可减少因螺母重合或遮挡导致识别率低的问题,并且它还增强了特征的传播和复用,可以提升不同天气条件和复杂环境下螺母检测的准确性。
YOLOv5默认使用的损失函数是CIoU(complete intersection-over-union),现更换为SIoU[17]。它在CIoU的基础上,解决了纵横比的模糊定义,并添加Focal Loss解决边界框回归(bounding boxes regression,BBox)中的样本不平衡问题,同时加入了类别信息的权重因子,加快了损失函数的收敛速度并提高了检测模型的分类准确率。SIoU还可以让预测框移动到最近的轴,然后回归一个坐标,这样能够减少自由度的总数,提高模型的回归精度。
SIoU进一步考虑了真实框和预测框之间的向量角度,重新定义相关损失函数,具体包含3个部分:角度损失(angle cost)、距离损失(distance cost)和形状损失(shape cost)[18]
1)角度损失
角度损失描述了中心点连接与X轴和Y轴之间的最小角度,如图7所示。
如果α π 4,则收敛过程最小化α,否则最小化,表达式为
β= π 2-α
为了实现这一过程,引入了LF(loss function)组件,角度损失函数Λ定义如下。
$\Lambda=1-2 \sin ^{2}\left(\arcsin x-\frac{\pi}{4}\right)$
式(2)中:
x= c h σ=sinα
σ= ( b c x g t - b c y ) 2 + ( b c y g t - b c y ) 2
ch=max( b c y g t, b c y)-min( b c y g t, b c y)
式中:(b c x, b c y)为预测框的中心;(b c x g t, b c y g t)为真实框的中心。
2)距离损失
距离损失描述了中心点之间的距离,距离损失函数Δ定义如下。
Δ= t = x , y (1-e-γρt)
式(6)中:
ρ x = b c x g t - b c x c w 2 ρ y = b c y g t - b c y c h 2
γ=2-Λ
可得,当α→0时,距离损失对整体损失的贡献很小。当α越接近 π 4,距离损失对整体损失的贡献越大。
3)形状损失
形状损失考虑两框之间的长宽比,通过计算两框之间长(宽)之差与二者之间最大长(宽)之比来定义的,θ对形状损失影响很大,其范围为[2,6],形状损失函数Ω定义如下。
Ω= t = w , h ( 1 - e - ω t ) θ
式(9)中:
ω w = w - w g t m a x ( w , w g t ) ω h = h - h g t m a x ( h , h g t )
形状损失图如图8所示。
最终SIoU损失函数为
Lbox=1-IoU+ Δ + Ω 2
式(11)中:
IoU= B B G T B B G T
实验在Windows11操作系统下进行,具体实验硬件及软件环境如表1所示。
为了更好地评估改进后模型的性能,采用准确率(precision, P)、召回率(recall, R)、平均精度均值1(mAP_0.5)(mean average precision)和平均精度均值2(mAP_0.5:0.95)作为模型性能的评价指标[19]。其中平均精度均值1指的是交并比(intersection-over-union, IoU)设置为0.5时,计算图片中每一类的AP(average precision),再将所有类别取平均值,即得到mAP_0.5。平均精度均值2指的是在IoU阈值从0.5到0.95、步长0.05上的平均mAP[20]
以YOLOv5模型为基础,在主干网络中添加高效多尺度注意力机制EMA,在颈部网络中更换PANet为加权特征融合的BiFPN,更换损失函数为SIoU,为了验证3种改进方法对模型实际性能的影响,进行了消融实验[21],各参数曲线图如图9所示,实验结果统计如表2所示。
表2中可得,实验1使用YOLOv5原模型作为实验基准,实验2~实验4对模型进行了单独改进,可以看出3个改进单独使用时参数有的提升,有的下降,但总体处于提升状态。实验5~实验7将3个改进进行两两组合,实验8将3个改进同时加入,相对于前面的实验,实验8参数提升最大,其中准确率提升了0.92%,召回率提升了0.16%,平均精度1提升了0.53%,平均精度2提升了2.26%;将该改进后的模型简称为YOLOv5-EBS模型。通过消融实验证明了改进方法对YOLOv5模型的性能具有积极的提升。
为了验证改进后的模型在实际检测过程中的效果,选取了一些遮挡、附着水泥、暗光条件下的螺母图像进行检测对比,效果如图10所示,在有遮挡导致螺母不完整的情况下YOLOv5原模型出现了漏检的情况,并且还出现了误检的情况,如图10(c)把螺母(标签1)识别成了螺栓(标签2),而改进后的YOLOv5-EBS模型则具有较好的检测表现。在螺母附着水泥时YOLOv5原模型出现了多次识别的情况,将螺母识别成了螺母和螺栓,如图10(e)的螺母上有两个识别框,而改进后的YOLOv5-EBS模型则没有出现该情况。
综上所述,在YOLOv5原模型基础上改进的YOLOv5-EMA-BiFPN-SIoU模型在塔吊螺母的实际识别过程中可以达到更好的识别效果,提高了螺母在遮挡情况下的识别率,并降低了天气等环境因素造成的漏检和误检的情况。
基于目标检测模型YOLOv5进行多项改进,以提升高空螺母的检测率。进行实验后得到以下结论。
(1)消融实验表明,YOLOv5改进后mAP_0.5提升了0.53%,mAP_0.5:0.95提升了2.26%。各项性能参数均有提升,比原模型具有更好的识别性能。
(2)实际识别对比实验表明,改进后的模型提高了被遮挡螺母的识别率,降低了螺母的误检、漏检率,为后续的螺母检测提供了良好的技术支持。
  • 国家自然科学基金(61973290)
参考文献 引证文献
排序方式:
[1]
黄志成, 吴永良, 梁臣. 基于BIM技术的格构式塔吊基础设计应用[J]. 中国建设信息化, 2023(20): 90-94.
Huang Zhicheng, Wu Yongliang, Liang Chen. Design and application of lattice tower crane foundation based on BIM technology[J]. China Construction Informatization, 2023(20): 90-94.
[2]
张知田, 王园园, 罗柱邦, 等. 塔吊与工人空间交互下危险场景自动检测[J]. 清华大学学报(自然科学版), 2024, 64 (2): 198-204.
Zhang Zhitian, Wang Yuanyuan, Luo Zhubang, et al. Automatic detection of hazardous scenes under the interaction between tower cranes and worker spaces[J]. Journal of Tsinghua University (Natural Science Edition), 2024, 64 (2): 198-204.
[3]
贾世娜. 基于改进YOLOv5的小目标检测算法研究[D]. 南昌: 南昌大学, 2022.
Jia Shina. Research on small object detection algorithm based on improved YOLOv5[D]. Nanchang: Nanchang University, 2022.
[4]
Yu F, Cai H, Zhang H, et al. Automatic identification method for three-phase structure of pervious concrete based on deep learning network of Mask R-CNN[J]. Construction and Building Materials, 2024, 420(1): 135534.
[5]
Zhang Y. Research on fabric yarn detection based on improved Fast R-CNN algorithm[J]. Applied Mathematics and Nonlinear Sciences, 2024, 9(1): 00449.
[6]
Yang Y. Vehicle target detection algorithm based on improved Faster R-CNN for remote sensing images[J]. Journal of Artificial Intelligence Practice, 2024, 7(1): 070105.
[7]
SangHyun L, Gao G. A study on pine larva detection system using swin transformer and cascade R-CNN hybrid model[J]. Applied Sciences, 2023, 13(3): 13031330.
[8]
Nguyen T N, Thanh Q D B, Tran C N N, et al. Improved detection network model based on YOLOv5 for warning safety in construction sites[J]. International Journal of Construction Management, 2024, 24(9): 1007-1017.
[9]
王景鑫, 潘欣. 一种基于Labelimg的辅助标注方法[J]. 科技创新与应用, 2023, 13(29): 145-148.
Wang Jingxin, Pan Xin. A labelimg based auxiliary annotation method[J]. Technological Innovation and Application, 2023, 13(29): 145-148.
[10]
Zou Z X, Shi Z W, Guo Y H, et al. Object detection in 20 years: a survey[J]. ArXiv, 2019: 1905.05055.
[11]
马永康, 刘华, 凌成星, 等. 基于改进YOLOv5的红树林单木目标检测研究[J]. 激光与光电子学进展, 2022, 59(18): 436-446.
Ma Yongkang, Liu Hua, Ling Chengxing, et al. Research on mangrove single tree target detection based on improved YOLOv5[J]. Progress in Laser and Optoelectronics, 2022, 59 (18): 436-446.
[12]
Hou Q, Zhou D, Feng J. Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition ( CVPR).Coordinate attention for efficient mobile network design[C]// Piscataway: IEEE, 2021: 13708-13717.
[13]
Ouyang D, He S, Zhang G, et al. Efficient multi-scale attention module with cross-spatial learning[C]// Proceedings of the ICASSP2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Rhodes: IEEE, 2023: 1-5.
[14]
李韬, 任玲, 胡斌, 等. 改进YOLOv5s和迁移学习对番茄穴盘苗的分级检测[J]. 农业工程学报, 2023, 39(23): 174-184.
Li Tao, Ren Ling, Hu Bin, et al. Improving YOLOv5s and transfer learning for grading detection of tomato hole seedlings[J]. Journal of Agricultural Engineering, 2023, 39(23): 174-184.
[15]
陈雷平, 段意强, 瞿宏愿. 基于改进YOLOv5的输电线路缺陷检测方法[J]. 技术与市场, 2024, 31(1): 54-56, 62.
Chen Leiping, Duan Yiqiang, Qu Hongyuan. A defect detection method for transmission lines based on improved YOLOv5[J]. Technology and Market, 2024, 31(1): 54-56, 62.
[16]
Tan M, Pang R, Le Q. V. Efficientdet:scalable and efficient object detection[C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition ( CVPR). Piscataway: IEEE, 2019: 10778-10787.
[17]
Gevorgyan Z. SIoU loss: more powerful learning for bounding box regression[J]. ArXiv, 2022: 2205.12740.
[18]
董恒祥, 潘江如, 董芙楠, 等. 基于YOLOv5s模型的边界框回归损失函数研究[J]. 现代电子技术, 2024, 47 (3): 179-186.
Dong Hengxiang, Pan Jiangru, Dong Funan, et al. Research on bounding box regression loss function based on YOLOv5s model[J]. Modern Electronic Technology, 2024, 47(3): 179-186.
[19]
孙俊, 吴兆祺, 贾忆琳, 等. 基于改进YOLOv5s的果园环境葡萄检测[J]. 农业工程学报, 2023, 39(18): 192-200.
Sun Jun, Wu Zhaoqi, Jia Yilin, et al. Grape detection in orchard environment based on improved YOLOv5s[J]. Journal of Agricultural Engineering, 2023, 39(18): 192-200.
[20]
张志远, 罗铭毅, 郭树欣, 等. 基于改进YOLOv5的自然环境下樱桃果实识别方法[J]. 农业机械学报, 2022, 53(S1): 232-240.
Zhang Zhiyuan, Luo Mingyi, Guo Shuxin, et al. Cherry fruit recognition method in natural environment based on improved YOLOv5[J]. Journal of Agricultural Machinery, 2022, 53(S1): 232-240.
[21]
赵桂平, 邓飞, 王昀, 等. 改进的YOLOv5-ResNet相似目标检测方法[J]. 科学技术与工程, 2022, 22(30): 13406-13416.
Zhao Guiping, Deng Fei, Wang Yun, et al. Improved YOLOv5 ResNet similar object detection method[J]. Science Technology and Engineering, 2022, 22(30): 13406-13416.
2025年第25卷第1期
PDF下载
349
144
引用本文
BibTeX
文章信息
doi: 10.12404/j.issn.1671-1815.2402469
  • 接收时间:2024-04-07
  • 首发时间:2025-07-29
  • 出版时间:2025-01-08
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-04-07
  • 修回日期:2024-10-15
基金
国家自然科学基金(61973290)
作者信息
    1.合肥大学先进制造工程学院, 合肥 230601
    2.中科院合肥物质科学研究院智能机械研究所, 合肥 230031

通讯作者:

* 方薇(1977—),女,汉族,安徽合肥人,博士,副研究员。研究方向:计算机信息处理、遥感技术应用。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2402469
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