Article(id=1251505537351893445, tenantId=1146029695717560320, journalId=1251233954884272221, issueId=1251505536634667461, articleNumber=null, orderNo=null, doi=10.13682/j.issn.2095-6533.2025.06.014, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1732723200000, receivedDateStr=2024-11-28, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1776311771953, onlineDateStr=2026-04-16, pubDate=1762704000000, pubDateStr=2025-11-10, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1776311771953, onlineIssueDateStr=2026-04-16, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1776311771953, creator=13701087609, updateTime=1776311771953, updator=13701087609, issue=Issue{id=1251505536634667461, tenantId=1146029695717560320, journalId=1251233954884272221, year='2025', volume='30', issue='6', pageStart='1', pageEnd='130', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1776311771782, creator=13701087609, updateTime=1776311824541, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1251505758014226723, tenantId=1146029695717560320, journalId=1251233954884272221, issueId=1251505536634667461, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1251505758014226724, tenantId=1146029695717560320, journalId=1251233954884272221, issueId=1251505536634667461, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=123, endPage=130, ext={EN=ArticleExt(id=1251505537557414345, articleId=1251505537351893445, tenantId=1146029695717560320, journalId=1251233954884272221, language=EN, title=Improved YOLOv8n surface defect detection algorithm for latex gloves, columnId=null, journalTitle=Journal of Xi'an University of Posts and Telecommunications, columnName=null, runingTitle=null, highlight=null, articleAbstract=

For the low detection accuracy problem of small targets and low-contrast defects on the surface of latex gloves,an improved YOLOv(you only look once version)8n algorithm for defects detection on the surface of latex gloves is proposed.The receptive field attention convolution module is introduced in the feature extraction network to dynamically adjust the spatial feature weights within the receptive field,and to enhance the network's focus on defect features. The C2f module is redesigned based on the proposed multi-scale convolution,which captures the contextual information from shallow features through multi-scale convolutional kernels,and improves the network's ability to extract shallow features.The context and the spatial feature calibration network are added to the feature fusion network,where feature calibration refines and aligns contextual information and spatial features,and further enhances the representation of defect features.Experimental results show that on the homemade dataset,the mean average precision(mAP)of the improved algorithm reaches 93.2%,which is 3.1%higher than that of YOLOv8n.It effectively improves the surface defect detection accuracy of latex gloves.In addition,on the VisDrone2019Det and steel defect detection datasets,the mAP reaches 36.1%and 79.8%,respectively,which are 1.1%and 2.7%higher than that of YOLOv8n,and further verifies the effectiveness of the improved algorithm.

, 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=Chunmei WANG, Guanying REN), CN=ArticleExt(id=1251505540061413910, articleId=1251505537351893445, tenantId=1146029695717560320, journalId=1251233954884272221, language=CN, title=改进YOLOv8n的乳胶手套表面缺陷检测算法, columnId=1251505537641300427, journalTitle=西安邮电大学学报, columnName=人工智能目标检测, runingTitle=null, highlight=null, articleAbstract=

针对乳胶手套表面小目标和低对比度缺陷检测精度低的问题,提出一种改进YOLOv(You Only Look Once version)8n的乳胶手套表面缺陷检测算法。在特征提取网络中引入感受野注意力卷积模块,通过动态调整感受野内的空间特征权重,增强网络对缺陷特征的关注能力。基于多尺度卷积对C2f模块进行重设计,通过多尺度卷积核捕获浅层特征的上下文信息,提升网络对浅层特征的提取能力。最后,在特征融合网络中添加上下文和空间特征校准网络,通过特征校准实现对上下文信息和空间特征的细化与对齐,进一步增强缺陷特征的表达能力。实验结果表明,在自制数据集上,改进算法的平均精度均值(mean Average Precision,mAP)达到了93.2%,较YOLOv8n提升了3.1%,有效提高了乳胶手套表面缺陷检测精度。另外,在VisDrone2019Det和钢铁缺陷检测数据集上的mAP分别达到了36.1%和79.8%,较YOLOv8n分别提升了1.1%和2.7%,进一步验证了改进算法的有效性。

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=dXALwDJ+Nx1ThYgzXR4fCA==, magXml=Dgapli0psOND4v0NItCUOA==, pdfUrl=null, pdf=xhoQwFa9YpQTIIYOeuVMig==, pdfFileSize=3547399, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=P6cIHpbIDX/IdYR/fI8BNw==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=M/iKYXW+1Eb3iMYGJUoMrA==, mapNumber=null, authorCompany=null, fund=null, authors=

王春梅(1979-),女,甘肃白银人,硕士,西安邮电大学副教授,主要研究方向为智能算法、目标检测和软件测试等。E-mail:

任冠莹(1999-),女,河南洛阳人,西安邮电大学硕士研究生,主要研究方向为目标检测。E-mail:

, authorsList=王春梅, 任冠莹)}, authors=[Author(id=1251505540489232941, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wangcm0225@126.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1251505540573119028, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, authorId=1251505540489232941, language=EN, stringName=Chunmei WANG, firstName=Chunmei, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.School of Computer Science,Xi'an University of Posts and Telecommunications,Xi'an 710121,China
2.Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing,Xi'an 710121,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1251505540648616504, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, authorId=1251505540489232941, language=CN, stringName=王春梅, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.西安邮电大学计算机学院,陕西西安 710121
2.陕西省网络数据分析与智能处理重点实验室,陕西西安 710121, bio={"img":"9MldhD453xDZZSPfXPmdrQ==","content":"

王春梅(1979-),女,甘肃白银人,硕士,西安邮电大学副教授,主要研究方向为智能算法、目标检测和软件测试等。E-mail:

"}, bioImg=9MldhD453xDZZSPfXPmdrQ==, bioContent=

王春梅(1979-),女,甘肃白银人,硕士,西安邮电大学副教授,主要研究方向为智能算法、目标检测和软件测试等。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1251505540287906337, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, xref=1., ext=[AuthorCompanyExt(id=1251505540304683553, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540287906337, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Computer Science,Xi'an University of Posts and Telecommunications,Xi'an 710121,China), AuthorCompanyExt(id=1251505540313072162, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540287906337, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.西安邮电大学计算机学院,陕西西安 710121)]), AuthorCompany(id=1251505540388569637, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, xref=2., ext=[AuthorCompanyExt(id=1251505540392763942, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540388569637, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing,Xi'an 710121,China), AuthorCompanyExt(id=1251505540401152551, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540388569637, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.陕西省网络数据分析与智能处理重点实验室,陕西西安 710121)])]), Author(id=1251505540740891196, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=ying191027@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1251505540849943108, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, authorId=1251505540740891196, language=EN, stringName=Guanying REN, firstName=Guanying, middleName=null, lastName=REN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.School of Computer Science,Xi'an University of Posts and Telecommunications,Xi'an 710121,China
2.Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing,Xi'an 710121,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1251505540921246280, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, authorId=1251505540740891196, language=CN, stringName=任冠莹, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, address=1.西安邮电大学计算机学院,陕西西安 710121
2.陕西省网络数据分析与智能处理重点实验室,陕西西安 710121, bio={"img":"VcsqcoDD+p1viOiIRk8Wbg==","content":"

任冠莹(1999-),女,河南洛阳人,西安邮电大学硕士研究生,主要研究方向为目标检测。E-mail:

"}, bioImg=VcsqcoDD+p1viOiIRk8Wbg==, bioContent=

任冠莹(1999-),女,河南洛阳人,西安邮电大学硕士研究生,主要研究方向为目标检测。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1251505540287906337, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, xref=1., ext=[AuthorCompanyExt(id=1251505540304683553, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540287906337, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Computer Science,Xi'an University of Posts and Telecommunications,Xi'an 710121,China), AuthorCompanyExt(id=1251505540313072162, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540287906337, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.西安邮电大学计算机学院,陕西西安 710121)]), AuthorCompany(id=1251505540388569637, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, xref=2., ext=[AuthorCompanyExt(id=1251505540392763942, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540388569637, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing,Xi'an 710121,China), AuthorCompanyExt(id=1251505540401152551, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540388569637, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.陕西省网络数据分析与智能处理重点实验室,陕西西安 710121)])])], keywords=[Keyword(id=1251505541047075407, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, orderNo=1, keyword=small object detection), Keyword(id=1251505541156127319, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, orderNo=2, keyword=defect detection), Keyword(id=1251505541219041884, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, orderNo=3, keyword=YOLOv8n), Keyword(id=1251505541281956448, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, orderNo=4, keyword=feature enhancement), Keyword(id=1251505541336482402, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, orderNo=5, keyword=multi-scale convolution), Keyword(id=1251505541411979877, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, orderNo=1, keyword=小目标检测), Keyword(id=1251505541474894441, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, orderNo=2, keyword=缺陷检测), Keyword(id=1251505541550391917, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, orderNo=3, keyword=YOLOv8n), Keyword(id=1251505541609112176, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, orderNo=4, keyword=特征增强), Keyword(id=1251505541697192563, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, orderNo=5, keyword=多尺度卷积)], refs=[Reference(id=1251505543555269331, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=36, issue=6, pageStart=115, pageEnd=123, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=王伟, 万晓刚, journalName=西安工程大学学报, refType=null, unstructuredReference=王伟,万晓刚.结合注意力机制和特征融合的小目标检测方法[J].西安工程大学学报,2022,36(6):115-123., articleTitle=结合注意力机制和特征融合的小目标检测方法, refAbstract=null), Reference(id=1251505543622378200, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=36, issue=6, pageStart=115, pageEnd=123, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=WANG W, WAN X G, journalName=Journal of Xi'an Polytechnic University, refType=null, unstructuredReference=WANG W,WAN X G.Asmall target detection method combining attention mechanism and feature fusion[J].Journal of Xi'an Polytechnic University,2022, 36(6):115-123.(in Chinese), articleTitle=Asmall target detection method combining attention mechanism and feature fusion, refAbstract=null), Reference(id=1251505543697875676, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2020, volume=396, issue=null, pageStart=39, pageEnd=64, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=WU X W, SAHOO D, HOI S C H, journalName=Neurocomputing, refType=null, unstructuredReference=WU X W,SAHOO D,HOI S C H.Recent advances in deep learning for object detection[J].Neurocomputing,2020,396:39-64., articleTitle=Recent advances in deep learning for object detection, refAbstract=null), Reference(id=1251505543802733282, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2021, volume=11, issue=2, pageStart=576, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=ZHANG K H, SHEN H K, journalName=Applied Sciences, refType=null, unstructuredReference=ZHANG K H,SHEN H K.Solder joint defect detection in the connectors using improved faster-RCNN algorithm[J].Applied Sciences,2021,11(2):576., articleTitle=Solder joint defect detection in the connectors using improved faster-RCNN algorithm, refAbstract=null), Reference(id=1251505543924368104, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=61, issue=11, pageStart=8963, pageEnd=8971, url=null, language=null, rfNumber=[4], rfOrder=4, authorNames=YANG A M, JIANG T Y, HAN Y, journalName=Alexandria Engineering Journal, refType=null, unstructuredReference=YANG A M,JIANG T Y,HAN Y,et al.Research on application of on-line melting in-situ visual inspection of iron ore powder based on faster R-CNN[J].Alexandria Engineering Journal, 2022, 61 (11):8963-8971., articleTitle=Research on application of on-line melting in-situ visual inspection of iron ore powder based on faster R-CNN, refAbstract=null), Reference(id=1251505543991476971, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=173, pageEnd=181, url=null, language=null, rfNumber=[5], rfOrder=5, authorNames=KUMAR A, MANIKANDAN R, journalName=null, refType=null, unstructuredReference=KUMAR A,MANIKANDAN R.Brain tumor detection using deep neural network-based classifier[C]//International Conference on Innovative Computing and Communications. Singapore: Springer, 2022:173-181., articleTitle=Brain tumor detection using deep neural network-based classifier, refAbstract=null), Reference(id=1251505544058585838, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=710, pageEnd=715, url=null, language=null, rfNumber=[6], rfOrder=6, authorNames=LIU T, CHEN S Y, journalName=null, refType=null, unstructuredReference=LIU T,CHEN S Y.YOLOv4-DCN-based fabric defect detection algorithm[C]//2022 37th Youth Academic Annual Conference of Chinese Association of Automation.Beijing:IEEE,2023:710-715., articleTitle=YOLOv4-DCN-based fabric defect detection algorithm, refAbstract=null), Reference(id=1251505544117306100, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=199, issue=null, pageStart=1066, pageEnd=1073, url=null, language=null, rfNumber=[7], rfOrder=7, authorNames=JIANG P Y, ERGU D, LIU F Y, journalName=Procedia Computer Science, refType=null, unstructuredReference=JIANG P Y,ERGU D,LIU F Y,et al.A review of yolo algorithm developments[J].Procedia Computer Science,2022,199:1066-1073., articleTitle=A review of yolo algorithm developments, refAbstract=null), Reference(id=1251505544196997878, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2023, volume=28, issue=3, pageStart=88, pageEnd=96, url=null, language=null, rfNumber=[8], rfOrder=8, authorNames=江祥奎, 杜遥遥, 胡浩昌, journalName=西安邮电大学学报, refType=null, unstructuredReference=江祥奎,杜遥遥,胡浩昌.一种改进YOLOv5s小目标无人机实时检测算法[J].西安邮电大学学报,2023, 28(3):88-96., articleTitle=一种改进YOLOv5s小目标无人机实时检测算法, refAbstract=null), Reference(id=1251505544289272572, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2023, volume=28, issue=3, pageStart=88, pageEnd=96, url=null, language=null, rfNumber=[8], rfOrder=9, authorNames=JIANG X K, DU Y Y, HU H C, journalName=Journal of Xi'an University of Posts and Telecommunications, refType=null, unstructuredReference=JIANG X K, DU Y Y, HU H C. An improved YOLOv5s real-time detection algorithm for small target UAVs[J].Journal of Xi'an University of Posts and Telecommunications,2023,28(3):88-96.(in Chinese), articleTitle=An improved YOLOv5s real-time detection algorithm for small target UAVs, refAbstract=null), Reference(id=1251505544364770048, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2023, volume=12, issue=13, pageStart=2821, pageEnd=null, url=null, language=null, rfNumber=[9], rfOrder=10, authorNames=DU B W, WAN F, LEI G B, journalName=Electronics, refType=null, unstructuredReference=DU B W,WAN F,LEI G B,et al.YOLO-MBBi:PCB surface defect detection method based on enhanced YOLOv5[J].Electronics,2023,12(13):2821., articleTitle=YOLO-MBBi:PCB surface defect detection method based on enhanced YOLOv5, refAbstract=null), Reference(id=1251505544452850440, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=71, issue=null, pageStart=2510011, pageEnd=null, url=null, language=null, rfNumber=[10], rfOrder=11, authorNames=YEUNG C C, LAM K M, journalName=IEEE Transactions on Instrumentation and Measurement, refType=null, unstructuredReference=YEUNG C C,LAM K M.Efficient fused-attention model for steel surface defect detection[J]. IEEE Transactions on Instrumentation and Measurement, 2022,71:2510011., articleTitle=Efficient fused-attention model for steel surface defect detection, refAbstract=null), Reference(id=1251505544528347914, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=478, pageEnd=483, url=null, language=null, rfNumber=[11], rfOrder=12, authorNames=REN Y H, ZHANG T, MA R H, journalName=null, refType=null, unstructuredReference=REN Y H,ZHANG T,MA R H,et al.Insulator defect detection method based on improved YOLOv7[C]//20239th International Conference on Mechanical and Electronics Engineering. Xi'an: IEEE, 2024:478-483., articleTitle=Insulator defect detection method based on improved YOLOv7, refAbstract=null), Reference(id=1251505544603845389, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=1192, pageEnd=1197, url=null, language=null, rfNumber=[12], rfOrder=13, authorNames=WANG H B, WANG Y X, journalName=null, refType=null, unstructuredReference=WANG H B,WANG Y X.Improved glove defect detection algorithm based on YOLOv5 framework[C]//2022 IEEE 6th Advanced Information Technology,Electronic and Automation Control Conference.Beijing:IEEE,2022:1192-1197., articleTitle=Improved glove defect detection algorithm based on YOLOv5 framework, refAbstract=null), Reference(id=1251505544670954258, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2023, volume=13, issue=18, pageStart=10173, pageEnd=null, url=null, language=null, rfNumber=[13], rfOrder=14, authorNames=JIN H L, DU R Y, QIAO L Y, journalName=Applied Sciences, refType=null, unstructuredReference=JIN H L,DU R Y,QIAO L Y,et al.CCA-YOLO:An improved glove defect detection algorithm based on YOLOv5[J].Applied Sciences,2023,13(18):10173., articleTitle=CCA-YOLO:An improved glove defect detection algorithm based on YOLOv5, refAbstract=null), Reference(id=1251505544771617559, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=5, issue=2, pageStart=36, pageEnd=53, url=null, language=null, rfNumber=[14], rfOrder=15, authorNames=CHU Z, WENG G, YU L, journalName=Artificial Intelligence and Machine Learning Review, refType=null, unstructuredReference=CHU Z,WENG G,YU L.Real-time industrial surface defect detection based on lightweight convolutional neural networks[J].Artificial Intelligence and Machine Learning Review,2024,5(2):36-53., articleTitle=Real-time industrial surface defect detection based on lightweight convolutional neural networks, refAbstract=null), Reference(id=1251505544868086557, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=29, issue=3, pageStart=75, pageEnd=82, url=null, language=null, rfNumber=[15], rfOrder=16, authorNames=江祥奎, 杨刚, 杜遥遥, journalName=西安邮电大学学报, refType=null, unstructuredReference=江祥奎,杨刚,杜遥遥.基于轻量化YOLOv8n的动态视觉SLAM算法[J].西安邮电大学学报,2024,29 (3):75-82., articleTitle=基于轻量化YOLOv8n的动态视觉SLAM算法, refAbstract=null), Reference(id=1251505544947778336, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=29, issue=3, pageStart=75, pageEnd=82, url=null, language=null, rfNumber=[15], rfOrder=17, authorNames=JIANG X K, YANG G, DU Y Y, journalName=Journal of Xi'an University of Posts and Telecommunications, refType=null, unstructuredReference=JIANG X K,YANG G,DU Y Y.Dynamic visual SLAM algorithm based on lightweight YOLOv8n[J]. Journal of Xi'an University of Posts and Telecommunications,2024,29(3):75-82.(in Chinese), articleTitle=Dynamic visual SLAM algorithm based on lightweight YOLOv8n, refAbstract=null), Reference(id=1251505545056830246, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=32, issue=14, pageStart=19, pageEnd=24, url=null, language=null, rfNumber=[16], rfOrder=18, authorNames=赵旭东, 王馨, 刘兴, journalName=电子设计工程, refType=null, unstructuredReference=赵旭东,王馨,刘兴,.改进YOLOv5的拥挤场景目标检测算法[J].电子设计工程,2024,32(14):19-24., articleTitle=改进YOLOv5的拥挤场景目标检测算法, refAbstract=null), Reference(id=1251505545136522025, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=32, issue=14, pageStart=19, pageEnd=24, url=null, language=null, rfNumber=[16], rfOrder=19, authorNames=ZHAO X D, WANG X, LIU X, journalName=Electronic Design Engineering, refType=null, unstructuredReference=ZHAO X D,WANG X,LIU X,et al.Improving YOLOv5's crowded scene object detection algorithm[J].Electronic Design Engineering,2024,32(14):19-24.(in Chinese), articleTitle=Improving YOLOv5's crowded scene object detection algorithm, refAbstract=null), Reference(id=1251505545232991021, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=73, issue=null, pageStart=5012614, pageEnd=null, url=null, language=null, rfNumber=[17], rfOrder=20, authorNames=HUANG Z L, ZHANG C X, GE L Y, journalName=IEEE Transactions on Instrumentation and Measurement, refType=null, unstructuredReference=HUANG Z L,ZHANG C X,GE L Y,et al.Joining spatial deformable convolution and a dense feature pyramid for surface defect detection[J].IEEE Transactions on Instrumentation and Measurement,2024,73:5012614., articleTitle=Joining spatial deformable convolution and a dense feature pyramid for surface defect detection, refAbstract=null), Reference(id=1251505545325265713, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=18, issue=11, pageStart=8327, pageEnd=8335, url=null, language=null, rfNumber=[18], rfOrder=21, authorNames=SUN F X, HE N, WANG X, journalName=Signal,Image and Video Processing, refType=null, unstructuredReference=SUN F X,HE N,WANG X,et al.YOLOv7-P:A lighter and more effective UAV aerial photography object detection algorithm[J].Signal,Image and Video Processing,2024,18(11):8327-8335., articleTitle=YOLOv7-P:A lighter and more effective UAV aerial photography object detection algorithm, refAbstract=null), Reference(id=1251505545430123318, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=81, issue=1, pageStart=155, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=22, authorNames=WEI H, ZHAO L C, LI R P, journalName=The Journal of Supercomputing, refType=null, unstructuredReference=WEI H,ZHAO L C,LI R P,et al.RFAConv-CBM-ViT:Enhanced vision transformer for metal surface defect detection[J].The Journal of Supercomputing, 2024,81(1):155., articleTitle=RFAConv-CBM-ViT:Enhanced vision transformer for metal surface defect detection, refAbstract=null), Reference(id=1251505545555952446, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=5672, pageEnd=5683, url=null, language=null, rfNumber=[20], rfOrder=23, authorNames=YU W H, ZHOU P, YAN S C, journalName=null, refType=null, unstructuredReference=YU W H,ZHOU P,YAN S C,et al.InceptionNeXt:When inception meets ConvNeXt[C]//2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Seattle:IEEE,2024:5672-5683., articleTitle=InceptionNeXt:When inception meets ConvNeXt, refAbstract=null), Reference(id=1251505545618867011, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=14, issue=21, pageStart=5488, pageEnd=null, url=null, language=null, rfNumber=[21], rfOrder=24, authorNames=SHI T J, GONG J N, HU J M, journalName=Remote Sensing, refType=null, unstructuredReference=SHI T J,GONG J N,HU J M,et al.Feature-enhanced CenterNet for small object detection in remote sensing images[J].Remote Sensing,2022,14 (21):5488., articleTitle=Feature-enhanced CenterNet for small object detection in remote sensing images, refAbstract=null), Reference(id=1251505545711141705, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2022, volume=52, issue=9, pageStart=10430, pageEnd=10447, url=null, language=null, rfNumber=[22], rfOrder=25, authorNames=YAN C Q, ZHANG H, LI X L, journalName=Applied Intelligence, refType=null, unstructuredReference=YAN C Q,ZHANG H,LI X L,et al.R-SSD:Refined single shot multibox detector for pedestrian detection[J]. Applied Intelligence, 2022, 52 (9):10430-10447., articleTitle=R-SSD:Refined single shot multibox detector for pedestrian detection, refAbstract=null), Reference(id=1251505545786639182, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2021, volume=10, issue=7, pageStart=837, pageEnd=null, url=null, language=null, rfNumber=[23], rfOrder=26, authorNames=JIANG X B, GAO T H, ZHU Z C, journalName=Electronics, refType=null, unstructuredReference=JIANG X B,GAO T H,ZHU Z C,et al.Real-time face mask detection method based on YOLOv3[J].Electronics,2021,10(7):837., articleTitle=Real-time face mask detection method based on YOLOv3, refAbstract=null), Reference(id=1251505545857942353, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=7464, pageEnd=7475, url=null, language=null, rfNumber=[24], rfOrder=27, authorNames=WANG C Y, BOCHKOVSKIY A, LIAO H M, journalName=null, refType=null, unstructuredReference=WANG C Y, BOCHKOVSKIY A, LIAO H M. YOLOv7:Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Vancouver:IEEE,2023:7464-7475., articleTitle=YOLOv7:Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors, refAbstract=null), Reference(id=1251505545971188564, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2025, volume=null, issue=null, pageStart=1, pageEnd=21, url=null, language=null, rfNumber=[25], rfOrder=28, authorNames=WANG C Y, YEH I H, MARK LIAO H Y, journalName=null, refType=null, unstructuredReference=WANG C Y, YEH I H, MARK LIAO H Y. YOLOv9:Learning what you want to learn using programmable gradient information[C]//Computer Vision-ECCV 2024.Cham:Springer,2025:1-21., articleTitle=YOLOv9:Learning what you want to learn using programmable gradient information, refAbstract=null), Reference(id=1251505546067657563, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, doi=null, pmid=null, pmcid=null, year=2025, volume=37, issue=null, pageStart=107984, pageEnd=108011, url=null, language=null, rfNumber=[26], rfOrder=29, authorNames=WANG A, CHEN H, LIU L, journalName=Advances in Neural Information Processing Systems, refType=null, unstructuredReference=WANG A,CHEN H,LIU L,et al.YOLOv10:Real-time end-to-end object detection[J].Advances in Neural Information Processing Systems, 2025, 37:107984-108011., articleTitle=YOLOv10:Real-time end-to-end object detection, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1251505540287906337, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, xref=1., ext=[AuthorCompanyExt(id=1251505540304683553, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540287906337, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.School of Computer Science,Xi'an University of Posts and Telecommunications,Xi'an 710121,China), AuthorCompanyExt(id=1251505540313072162, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540287906337, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.西安邮电大学计算机学院,陕西西安 710121)]), AuthorCompany(id=1251505540388569637, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, xref=2., ext=[AuthorCompanyExt(id=1251505540392763942, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540388569637, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing,Xi'an 710121,China), AuthorCompanyExt(id=1251505540401152551, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, companyId=1251505540388569637, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.陕西省网络数据分析与智能处理重点实验室,陕西西安 710121)])], figs=[ArticleFig(id=1251505541823021690, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=NXfhOyQMhdkgQd+5sxn8EQ==, figureFileBig=P6cIHpbIDX/IdYR/fI8BNw==, tableContent=null), ArticleFig(id=1251505541902713471, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=图1, caption=改进算法的网络结构示意图, figureFileSmall=NXfhOyQMhdkgQd+5sxn8EQ==, figureFileBig=P6cIHpbIDX/IdYR/fI8BNw==, tableContent=null), ArticleFig(id=1251505542083068554, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=pPcOzcJGGZSCH4u3N58fWA==, figureFileBig=uCK1ku8tptMyAj2A3/M8Ow==, tableContent=null), ArticleFig(id=1251505542150177422, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=图2, caption=C2f_MSC结构示意图, figureFileSmall=pPcOzcJGGZSCH4u3N58fWA==, figureFileBig=uCK1ku8tptMyAj2A3/M8Ow==, tableContent=null), ArticleFig(id=1251505542238257810, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=/3Kajeoo1YHJKGyy3uNU4g==, figureFileBig=v5qEpbuW91XN6V9yPINu5g==, tableContent=null), ArticleFig(id=1251505542305366677, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=图3, caption=CFC模块结构示意图, figureFileSmall=/3Kajeoo1YHJKGyy3uNU4g==, figureFileBig=v5qEpbuW91XN6V9yPINu5g==, tableContent=null), ArticleFig(id=1251505542364086937, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=plAYXYA0t0mv7b+h1N6+rw==, figureFileBig=P6vuu6AOcKgGtW6A9MZD7w==, tableContent=null), ArticleFig(id=1251505542435390109, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=图4, caption=SFC模块结构示意图, figureFileSmall=plAYXYA0t0mv7b+h1N6+rw==, figureFileBig=P6vuu6AOcKgGtW6A9MZD7w==, tableContent=null), ArticleFig(id=1251505542536053412, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=Onyep8u8jiTEIS0pw0A5Ig==, figureFileBig=7CQ5v/rZN1dESP/JoJ4diQ==, tableContent=null), ArticleFig(id=1251505542670271147, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=图5, caption=乳胶手套8种缺陷检测情况, figureFileSmall=Onyep8u8jiTEIS0pw0A5Ig==, figureFileBig=7CQ5v/rZN1dESP/JoJ4diQ==, tableContent=null), ArticleFig(id=1251505542812877488, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=Jhzydukgm+TZA2lM7JoZeQ==, figureFileBig=V7k3dPzHqeqbqdCoRT8FAQ==, tableContent=null), ArticleFig(id=1251505542905152179, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=图6, caption=两种算法在GLO-DET数据集上的检测结果, figureFileSmall=Jhzydukgm+TZA2lM7JoZeQ==, figureFileBig=V7k3dPzHqeqbqdCoRT8FAQ==, tableContent=null), ArticleFig(id=1251505542980649655, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
实验ABCAP/%mAP/%Params/MGFLOPS
胶粒误印斑点锈迹昆虫杂质洞粘褶针孔
190.095.675.190.199.398.691.680.690.13.018.1
293.894.678.591.199.598.994.782.991.83.038.4
394.894.479.793.198.898.994.082.692.02.727.6
488.695.278.791.598.899.394.380.990.93.328.9
593.494.982.393.999.598.992.985.492.62.747.9
692.595.381.695.499.598.994.387.893.23.068.8
), ArticleFig(id=1251505543068730043, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=表1, caption=

GLO-DET数据集上的消融实验结果

, figureFileSmall=null, figureFileBig=null, tableContent=
实验ABCAP/%mAP/%Params/MGFLOPS
胶粒误印斑点锈迹昆虫杂质洞粘褶针孔
190.095.675.190.199.398.691.680.690.13.018.1
293.894.678.591.199.598.994.782.991.83.038.4
394.894.479.793.198.898.994.082.692.02.727.6
488.695.278.791.598.899.394.380.990.93.328.9
593.494.982.393.999.598.992.985.492.62.747.9
692.595.381.695.499.598.994.387.893.23.068.8
), ArticleFig(id=1251505543173587647, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
检测算法AP/%mAP/%Params/MGFLOPS
胶粒误印斑点锈迹昆虫杂质洞粘褶针孔
Faster RCNN82.193.757.984.498.581.287.374.682.541.53207.0
CenterNet84.293.969.681.895.490.679.673.983.644.6355.6
SSD84.994.657.385.599.099.088.878.686.126.3031.2
YOLOv343.380.442.159.391.984.231.869.362.361.5065.86
YOLOv5s94.594.482.687.999.299.591.682.091.47.2016.50
YOLOv794.492.779.785.599.399.590.577.889.936.94104.70
YOLOv8n90.095.675.190.199.398.691.680.690.13.018.10
YOLOv9c91.095.182.091.798.399.593.182.891.725.30102.40
YOLOv10n87.393.874.688.299.098.987.978.888.62.266.50
改进算法92.595.381.695.499.598.994.387.893.23.068.80
), ArticleFig(id=1251505543265862338, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=表2, caption=

不同检测算法的检测精度对比

, figureFileSmall=null, figureFileBig=null, tableContent=
检测算法AP/%mAP/%Params/MGFLOPS
胶粒误印斑点锈迹昆虫杂质洞粘褶针孔
Faster RCNN82.193.757.984.498.581.287.374.682.541.53207.0
CenterNet84.293.969.681.895.490.679.673.983.644.6355.6
SSD84.994.657.385.599.099.088.878.686.126.3031.2
YOLOv343.380.442.159.391.984.231.869.362.361.5065.86
YOLOv5s94.594.482.687.999.299.591.682.091.47.2016.50
YOLOv794.492.779.785.599.399.590.577.889.936.94104.70
YOLOv8n90.095.675.190.199.398.691.680.690.13.018.10
YOLOv9c91.095.182.091.798.399.593.182.891.725.30102.40
YOLOv10n87.393.874.688.299.098.987.978.888.62.266.50
改进算法92.595.381.695.499.598.994.387.893.23.068.80
), ArticleFig(id=1251505543370719944, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=EN, label=null, caption=null, figureFileSmall=null, figureFileBig=null, tableContent=
数据集检测算法P/%R/%mAP/%
VisDrone2019DetYOLOv8n46.734.635.0
改进算法47.436.236.1
NEU-DETYOLOv8n70.175.877.1
改进算法80.572.779.8
), ArticleFig(id=1251505543467188940, tenantId=1146029695717560320, journalId=1251233954884272221, articleId=1251505537351893445, language=CN, label=表3, caption=

两种算法在其他数据集上的实验结果

, figureFileSmall=null, figureFileBig=null, tableContent=
数据集检测算法P/%R/%mAP/%
VisDrone2019DetYOLOv8n46.734.635.0
改进算法47.436.236.1
NEU-DETYOLOv8n70.175.877.1
改进算法80.572.779.8
)], attaches=null, journal=Journal(id=1251231493423411294, delFlag=0, nameCn=西安邮电大学学报, nameEn=Journal of Xi'an University of Posts and Telecommunications, nameHistory1=null, nameHistory2=null, issn=2095-6533, eissn=, cn=61-1493/TN, coden=null, periodic=双月刊, language=CN, oaType=1, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=, officePhone=, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=tH6CrDBQFqwqmwUccjpjow==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Xi'an University of Posts and Telecommunications, journalRemark=null, publicationField=null, createdTime=1776246434792, updatedTime=1776251858439, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Natural Sciences, subjectName=自然科学, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=tH6CrDBQFqwqmwUccjpjow==, picEn=TGVaHCgPGQlZdesVp9aDew==, jcr=null, cjcr=null, exts=[JournalExt(id=1251254241990493156, language=CN, name=西安邮电大学学报, 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=1776251858470, updatedTime=1776251858470, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://xayd.cbpt.cnki.net/index.aspx?t=1, submissionEditorUrl=https://xayd.cbpt.cnki.net/index.aspx?t=3, submissionReviewUrl=https://xayd.cbpt.cnki.net/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1251254242036630501, language=EN, name=Journal of Xi'an University of Posts and Telecommunications, 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=1776251858481, updatedTime=1776251858481, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://xayd.cbpt.cnki.net/index.aspx?t=1, submissionEditorUrl=https://xayd.cbpt.cnki.net/index.aspx?t=3, submissionReviewUrl=https://xayd.cbpt.cnki.net/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1251233954884272221, websiteList=[Website(id=1251257283410346025, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1251233954884272221, 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/xayddxxb/CN, language=CN, createTime=1776252583601, createBy=18614031015, updateTime=1776252942185, updateBy=18614031015, name=西安邮电大学学报-中文, tplId=1146099689490845704, title=西安邮电大学学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1251258885185683591, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=articleTextType, value=kx, createTime=1776252965494, updateTime=1776252965494, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885160517764, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=banner, value=null, createTime=1776252965488, updateTime=1776252965488, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885202460810, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=grayFlag, value=0, createTime=1776252965498, updateTime=1776252965498, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885152129155, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=logo, value=https://castjournals.cast.org.cn/joweb/xayddxxb/CN/file/pic?fileId=oIRdMBF7r3ynDYM5hP49NA==, createTime=1776252965486, updateTime=1776252965486, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885219238028, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=minRunFlag, value=0, createTime=1776252965502, updateTime=1776252965502, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885177294982, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/xayddxxb/CN/file/pic, createTime=1776252965492, updateTime=1776252965492, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885210849419, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=silenceFlag, value=0, createTime=1776252965500, updateTime=1776252965500, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885168906373, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1776252965490, updateTime=1776252965490, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885194072200, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=themeColor, value=null, createTime=1776252965496, updateTime=1776252965496, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258885198266505, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283410346025, code=themeStyle, value=null, createTime=1776252965497, updateTime=1776252965497, creator=18614031015, updator=18614031015)]), Website(id=1251257283594895425, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1251233954884272221, 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/xayddxxb/EN, language=EN, createTime=1776252583645, createBy=18614031015, updateTime=1776252932924, updateBy=18614031015, name=西安邮电大学学报-英文, tplId=1146101810881728533, title=Journal of Xi'an University of Posts and Telecommunications, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1251258912788398232, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=articleTextType, value=kx, createTime=1776252972075, updateTime=1776252972075, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912767426709, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=banner, value=null, createTime=1776252972070, updateTime=1776252972070, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912805175451, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=grayFlag, value=0, createTime=1776252972079, updateTime=1776252972079, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912759038100, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=logo, value=https://castjournals.cast.org.cn/joweb/xayddxxb/EN/file/pic?fileId=oIRdMBF7r3ynDYM5hP49NA==, createTime=1776252972068, updateTime=1776252972068, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912813564061, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=minRunFlag, value=0, createTime=1776252972081, updateTime=1776252972081, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912780009623, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/xayddxxb/EN/file/pic, createTime=1776252972073, updateTime=1776252972073, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912809369756, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=silenceFlag, value=0, createTime=1776252972080, updateTime=1776252972080, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912771621014, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1776252972071, updateTime=1776252972071, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912792592537, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=themeColor, value=null, createTime=1776252972076, updateTime=1776252972076, creator=18614031015, updator=18614031015), WebsiteProps(id=1251258912796786842, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251257283594895425, code=themeStyle, value=null, createTime=1776252972077, updateTime=1776252972077, creator=18614031015, updator=18614031015)])], journalTitle=西安邮电大学学报, weixinUrl=null, journalUrl=https://xayd.cbpt.cnki.net/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Xi'an University of Posts and Telecommunications, journalPhotoCn=tH6CrDBQFqwqmwUccjpjow==, journalPhotoEn=TGVaHCgPGQlZdesVp9aDew==, journalFirstLetter=J, 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/xayddxxb/CN/10.13682/j.issn.2095-6533.2025.06.014, detailUrlEn=https://castjournals.cast.org.cn/joweb/xayddxxb/EN/10.13682/j.issn.2095-6533.2025.06.014, pdfUrlCn=https://castjournals.cast.org.cn/joweb/xayddxxb/CN/PDF/10.13682/j.issn.2095-6533.2025.06.014, pdfUrlEn=https://castjournals.cast.org.cn/joweb/xayddxxb/EN/PDF/10.13682/j.issn.2095-6533.2025.06.014, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
改进YOLOv8n的乳胶手套表面缺陷检测算法
收藏切换
PDF下载
王春梅 1, 2 , 任冠莹 1, 2
西安邮电大学学报 | 人工智能目标检测 2025,30(6): 123-130
收起
收藏切换
西安邮电大学学报 | 人工智能目标检测 2025, 30(6): 123-130
改进YOLOv8n的乳胶手套表面缺陷检测算法
全屏
王春梅1, 2 , 任冠莹1, 2
作者信息
  • 1.西安邮电大学计算机学院,陕西西安 710121
  • 2.陕西省网络数据分析与智能处理重点实验室,陕西西安 710121
  • 王春梅(1979-),女,甘肃白银人,硕士,西安邮电大学副教授,主要研究方向为智能算法、目标检测和软件测试等。E-mail:

    任冠莹(1999-),女,河南洛阳人,西安邮电大学硕士研究生,主要研究方向为目标检测。E-mail:

Improved YOLOv8n surface defect detection algorithm for latex gloves
Chunmei WANG1, 2 , Guanying REN1, 2
Affiliations
  • 1.School of Computer Science,Xi'an University of Posts and Telecommunications,Xi'an 710121,China
  • 2.Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing,Xi'an 710121,China
出版时间: 2025-11-10 doi: 10.13682/j.issn.2095-6533.2025.06.014
文章导航
收藏切换

针对乳胶手套表面小目标和低对比度缺陷检测精度低的问题,提出一种改进YOLOv(You Only Look Once version)8n的乳胶手套表面缺陷检测算法。在特征提取网络中引入感受野注意力卷积模块,通过动态调整感受野内的空间特征权重,增强网络对缺陷特征的关注能力。基于多尺度卷积对C2f模块进行重设计,通过多尺度卷积核捕获浅层特征的上下文信息,提升网络对浅层特征的提取能力。最后,在特征融合网络中添加上下文和空间特征校准网络,通过特征校准实现对上下文信息和空间特征的细化与对齐,进一步增强缺陷特征的表达能力。实验结果表明,在自制数据集上,改进算法的平均精度均值(mean Average Precision,mAP)达到了93.2%,较YOLOv8n提升了3.1%,有效提高了乳胶手套表面缺陷检测精度。另外,在VisDrone2019Det和钢铁缺陷检测数据集上的mAP分别达到了36.1%和79.8%,较YOLOv8n分别提升了1.1%和2.7%,进一步验证了改进算法的有效性。

小目标检测  /  缺陷检测  /  YOLOv8n  /  特征增强  /  多尺度卷积

For the low detection accuracy problem of small targets and low-contrast defects on the surface of latex gloves,an improved YOLOv(you only look once version)8n algorithm for defects detection on the surface of latex gloves is proposed.The receptive field attention convolution module is introduced in the feature extraction network to dynamically adjust the spatial feature weights within the receptive field,and to enhance the network's focus on defect features. The C2f module is redesigned based on the proposed multi-scale convolution,which captures the contextual information from shallow features through multi-scale convolutional kernels,and improves the network's ability to extract shallow features.The context and the spatial feature calibration network are added to the feature fusion network,where feature calibration refines and aligns contextual information and spatial features,and further enhances the representation of defect features.Experimental results show that on the homemade dataset,the mean average precision(mAP)of the improved algorithm reaches 93.2%,which is 3.1%higher than that of YOLOv8n.It effectively improves the surface defect detection accuracy of latex gloves.In addition,on the VisDrone2019Det and steel defect detection datasets,the mAP reaches 36.1%and 79.8%,respectively,which are 1.1%and 2.7%higher than that of YOLOv8n,and further verifies the effectiveness of the improved algorithm.

small object detection  /  defect detection  /  YOLOv8n  /  feature enhancement  /  multi-scale convolution
王春梅, 任冠莹. 改进YOLOv8n的乳胶手套表面缺陷检测算法. 西安邮电大学学报, 2025 , 30 (6) : 123 -130 . DOI: 10.13682/j.issn.2095-6533.2025.06.014
Chunmei WANG, Guanying REN. Improved YOLOv8n surface defect detection algorithm for latex gloves[J]. Journal of Xi'an University of Posts and Telecommunications, 2025 , 30 (6) : 123 -130 . DOI: 10.13682/j.issn.2095-6533.2025.06.014
目标检测作为计算机视觉的核心任务之一,旨在通过算法自动定位并识别图像中的目标对象。随着深度学习技术的快速发展,目标检测技术在自动驾驶、安防监控和工业质检等领域得到了广泛应用[1]。尤其在工业生产中,自动化的缺陷检测系统正逐渐取代传统的人工检测,大幅提升了生产效率和产品质量。
当前,基于深度学习的目标检测算法主要分为两阶段目标检测算法与单阶段目标检测算法两类[2]。以快速区域卷积神经网络[3-5](Faster Region-based Convolutional Neural Networks,Faster RCNN)为代表的两阶段目标检测算法,虽提升了检测精度,但由于冗余计算,难以满足工业实时检测的要求[6]。相比之下,YOLO(You Only Look Once)系列[7]的单阶段目标检测算法,因其速度与精度之间的良好平衡,已成为工业缺陷检测的主流选择。特别是YOLO系列的第五个版本YOLOv5[8]已被广泛应用于印刷电路板检测[9]、钢表面缺陷检测[10]和绝缘子检测[11]等工业产品表面缺陷检测领域。然而,在乳胶手套表面缺陷检测中,YOLOv5存在精度不足的问题。Wang等[12]对YOLOv5框架进行改进,通过引入幻影网络和通道协调注意力机制,有效提升了整体检测效果,但对小目标缺陷检测的精度仍显不足。Jin等[13]提出了一种基于YOLOv5改进的手套缺陷检测算法,通过增加小目标检测层和创新的通道协调注意机制进一步优化了检测性能,但是大幅增加了参数量和计算量,不利于模型在实际场景中进行部署[14]。相比之下,YOLOv8[15]继承了YOLOv5的优点,在精度、速度和模型轻量化方面均有显著提升,尤其适用于工业实时检测场景。然而,YOLOv8n在乳胶手套小目标缺陷和低对比度缺陷检测中,由于其浅层特征提取能力不足,缺乏有效的特征增强机制,导致小目标缺陷与低对比度缺陷的检测精度低,漏检率和误检率高[16]。此外,在复杂背景下,小目标缺陷容易与背景噪声混淆,进一步影响了检测性能。
为解决上述问题,本文拟提出一种改进YOLOv8n的乳胶手套表面缺陷检测算法。通过引入感受野注意力卷积(Receptive Field Attention Convolution,RFAConv)模块,针对不同尺寸的缺陷自适应地分配注意力权重,突出感受野滑块内各个特征的重要性,提升小目标缺陷检测精度。通过多尺度卷积(Multi-Scale Convolution,MSConv)模块对C2f结构进行重设计,使模型在提取多尺度高层特征的同时保留其低层特征,防止小目标和低对比度缺陷被视为噪声,并利用上下文和空间特征校准网络(Context and Spatial Feature Calibration Network,CSFCN)优化多尺度特征的融合,缓解上下文失配问题,从而进一步提高模型在复杂场景下对小目标缺陷的检测精度。最后,在自制的乳胶手套表面缺陷检测(Glove Defect Detection,GLO-DET)数据集、钢铁缺陷检测[17](Northeastern University Steel Defect Detection,NEU-DET)数据集和小目标检测数据集VisDrone2019Det[18]上进行实验,以验证改进算法的有效性。
改进YOLOv8n的乳胶手套表面缺陷检测算法的整个网络结构主要由特征提取网络、特征融合网络和预测网络3部分组成。在特征提取网络中,RFAConv模块通过细化不同尺寸缺陷特征,增强网络对感受野空间特征的关注能力,提升对小目标缺陷的检测精度。在特征提取网络和特征融合网络中,采用MSConv模块对通道数大于256的C2f结构进行改进,设计了C2f_MSC模块,以增强网络对不同尺度特征及低对比度特征的关注度,避免对小目标有用的特征被当作噪声去除。此外,为进一步提升特征融合效果,在特征融合网络中添加CSF-CN模块,解决特征融合过程中出现的上下文不匹配和空间特征对不齐的问题,从而提升检测精度。通过3个模块的紧密协作,使模型在小目标缺陷和低对比度缺陷检测方面表现出色,降低漏检率和误检率,提升整体检测精度。改进算法的网络结构示意图如图1所示。
受Zhang等[19]提出的感受野注意力卷积启发,在特征提取网络中采用RFAConv模块,通过组卷积提取感受野特征,增强网络对感受野空间特征的关注能力,提高小目标特征提取的精确性。同时,注意力机制的引入使卷积核参数能够进行自适应调整,为不同区域提供定制化的处理。
在RFAConv中,首先采用快速的3×3组卷积提取感受野空间特征,通过平均池化聚合每个感受野特征的全局信息,避免特征交互冲突。然后,使用1×1组卷积交互信息,充分利用卷积核参数。通过Softmax生成的注意力图和感受野空间特征进行加权运算和形状调整,确保滑块内的特征不重叠。最后,再经过3×3卷积提取感受野内的滑块特征信息,最终生成特征图的表达式为
式中:C3×3表示3×3卷积运算;Softmax(·)和Relu(·)表示激活函数;P(·)表示平均池化;gi×i表示大小为i×i的分组卷积;k表示卷积核的大小;Norm(·)表示归一化;Ar表示生成的注意力图;Fr表示感受野空间特征;X表示输入特征图,F表示最终生成的特征图。
在YOLOv8n中,C2f结构的堆叠容易导致信道信息的冗余,且固定的卷积核限制了网络的感受野,难以有效捕获多尺度和低对比度缺陷间的细微差别。因此,针对现有C2f结构存在的问题,引入MSConv模块,设计C2f_MSC模块。将不同的核大小混合在一个卷积运算中,使模型在提取深层特征的同时保留其浅层特征,提升多尺度缺陷和低对比度缺陷的检测精度。C2f_MSC结构示意图如图2所示。
在MSConv中,首先将输入特征图X分成4组,表示为X=[X1X2X3X4],每组的卷积滤波器仅处理对应的输入特征图,其中1/4通道进行3×3卷积操作,另1/4通道进行5×5卷积操作,生成的特征图为
剩余的1/2通道保留其原始信息,生成的特征图为
分组后的特征图通过不同的卷积核尺寸进行处理,并重新排列成多个特征组{Y1Y2Y3Y4},并通过拼接操作重新组合通道。最后,使用逐点卷积进行通道间信息整合,增强多尺度特征的表达能力,生成最终的输出特征图为
式中:C1×1(·)表示逐点卷积操作;C(·)表示拼接操作。
这种分组处理策略能够同时处理浅层特征和深层特征,有效提升多尺度目标特征的捕获能力,优化了信息流和特征提取效率,同时保持网络的表征能力,从而提升模型的整体性能。
为了进一步提高乳胶手套表面缺陷检测模型的性能,在特征融合部分添加CSFCN模块[20],通过特征校准优化上下文和空间特征,进一步增强对小目标缺陷的检测能力。CSFCN包含上下文特征校准(Context Feature Calibration,CFC)和空间特征校准(Spatial Feature Calibration,SFC)两个核心模块。
CFC采用级联金字塔池化模块高效捕获上下文,并根据像素上下文相似度聚合语义相关区域,从而提升对小目标缺陷的敏感度。给定特征X,首先通过1×1卷积生成降维特征Q,并利用级联金字塔池化模块来提取多尺度上下文Z。接着,计算像素上下文亲和度,将其与输入特征相乘,获得校准后的语义上下文E。然后,通过调整响应值生成细粒度的上下文,最终生成的输出特征图为
式中:{XiYiαiZj}∈C×1表示输入、输出、重新校准因子和上下文;i的范围为[1,2,…,N],N表示像素,N=H×WHW分别表示特征图的高度和宽度;M表示上下文的总数;f(·)表示计算特征之间亲和度的成对函数。
CFC模块结构示意图如图3所示。
为进一步弥补逐步下采样造成的空间细节损失,SFC模块沿通道维度将特征分为多组子特征,并通过可学习采样传播子特征以实现空间特征校准。SFC模块结构示意图如图4所示。
首先,通过卷积对低分辨率特征FL和高分辨率特征统一通道数,并采用双线性插值法进行上采样,将其FH与进行拼接。然后,通过卷积预测偏移图ΔL、ΔH和门控掩码βLβH,调整跨层级特征对齐。最后,通过逐元素求和生成校准后的特征图为
式中:U(·)表示双线性上采样函数;T(·)表示校准函数;WLWH均表示具有批归一化处理和Relu的卷积层。
实验采用单张图形处理器(NVIDIA RTX A5000 24G)进行训练和测试。基于Pytorch 1.13.1框架和Python 3.8.17构建模型,训练轮数为300,每轮输入32张640×640的图像。默认学习率为0.01,使用余弦退火算法进行更新调整。优化器为随机梯度下降算法,动量和权重衰减值分别为0.937和0.0005。随机种子设置为0,确保每次训练结果的一致性。检测头的类别、边界框和分布焦点损失权重分布设为0.5、7.5和1.5。使用非极大值抑制去除得分较低的检测框,交并比阈值设置为0.7,置信度阈值设置为0.25。
针对生产车间的实际缺陷样本,对数据进行采集工作,确保乳胶手套数据集缺陷的真实性。自制GLO-DET数据集共有8种缺陷:胶粒、误印、斑点、锈迹、昆虫、杂质洞、粘褶和针孔,缺陷检测情况如图5所示。
为应对工业领域中由于良品率提升导致的缺陷样本少的问题,通过翻转、裁剪和亮度调整等数据增强方法,减少模型对特定特征的过度依赖,增强模型的泛化能力。此外,还进行数据清洗以保证数据的准确性。最终,扩充后的数据集为3120张,并按8∶2比例划分为训练集和测试集。为了客观评估算法的性能,采用平均精度均值(Mean Average Precision,mAP)、参数量(Parameters,Params)、精确率(Precision,P)、召回率(Recall,R)及每秒浮点运算量(Giga Floating Point Operations,GFLOPs)作为评估指标。
为了验证RFAConv、C2f_MSC和CSFCN模块在基线模型YOLOv8n上的性能增益,在GLO-DET数据集上进行6组消融实验。其中A代表RFAConv模块,B代表C2f_MSC模块,C代表CSFCN模块。消融实验结果如表1所示。
表1可知,实验1中基线模型YOLOv8n的mAP为90.1%,对斑点和针孔缺陷的检测能力较低。实验2引入A模块后,胶粒、斑点和粘褶缺陷检测精度分别提高了3.8%、3.4和3.1%,mAP提升1.7%。实验3采用B模块后,mAP提高了1.9%,这是因为B模块中的MSConv卷积在减少参数量和计算量的同时,捕获多尺度特征的上下文信息,增强网络对小目标和低对比缺陷的表征能力。实验4引入C模块后,通过校准上下文和空间特征,精细化处理小目标缺陷的特征信息,进一步提升了小目标缺陷的检测能力。实验5组合模块A和B后可以发现,小目标缺陷(如斑点缺陷)和低对比度缺陷(如针孔缺陷)的检测精度均有显著提升,分别提升了7.2%和5.4%,mAP提升了2.6%。实验6同时添加模块A、B、C后,即改进算法的mAP较基线模型提升了3.1%,进一步提高了检测精度,充分证明了算法的有效性,符合乳胶手套表面小目标缺陷检测的要求。
为验证改进算法的优越性,在GLO-DET数据集上复现了Faster RCNN、CenterNet[21]、单阶段检测器[22](Single Shot Multibox Detector,SSD)、YOLOv3[23]、YOLOv5s、YOLOv7[24]、YOLOv8n、YOLOv9c[25]和YOLOv10n[26]等先进的目标检测算法,并与改进算法进行对比实验,结果如表2所示。
表2可知,两阶段算法Faster RCNN的性能较差,无法满足实时检测需求。YOLOv8之前的单阶段算法,虽然相较于两阶段算法在精度上有所提升,但在参数量和计算量上仍显不足,限制了算法在资源受限环境中的应用。YOLOv9虽然通过多尺度特征融合和复杂结构将mAP提升至91.7%,但因其网络结构复杂,导致参数量和计算量显著增加,难以满足工业缺陷检测对实时性的要求。而YOLOv10由于追求速度提升,通过对特征提取模块进行轻量化处理,将参数量和计算量分别降低至2.26MB和6.5 GFLOPS,但这种轻量化处理削弱了对小目标和低对比度缺陷特征的提取能力,导致检测精度下降至88.6%,在实际场景中不利于高精度缺陷检测。因此,综合考虑检测精度和速度之间的平衡,选用YOLOv8n作为基线模型。在参数量和计算量利于实时检测的情况下,改进算法整体检测精度最高,mAP达到了93.2%,在乳胶手套表面缺陷检测中具有明显优势。
为了更直观地说明改进算法的检测效果,分别使用YOLOv8n和改进算法在GLO-DET数据集上对乳胶手套表面缺陷进行检测,结果如图6所示。对比图6前三列可以发现改进算法有效解决了低对比度缺陷的问题。对比图6后三列,面对小目标缺陷出现在边角的情况,改进算法也能够对其进行检测。整体来看,改进算法不仅能够检测出YOLOv8n漏检的小目标和低对比度的缺陷,而且在检测精度上也优于基线模型。因此,改进算法在乳胶手套表面小目标和低对比度缺陷的检测中具有较高的可行性和优越性。
为充分验证改进算法的泛化性,将YOLOv8n和改进算法分别在小目标数据集VisDrone2019Det和缺陷检测数据集NEU-DET上进行实验,结果如表3所示。
表3可知,改进算法在两个数据集上的检测精度均优于基线模型。在VisDrone2019Det数据集上,改进算法的mAP增加了1.1%,在NEU-DET数据集上提升了2.7%,表明该模型在不同数据集上的有效性和良好的泛化能力。
针对乳胶手套表面缺陷小、背景与缺陷难以区分的问题,提出了一种改进YOLOv8n的缺陷检测算法。通过引入RFAConv模块,自适应调整卷积核参数,聚焦感受野内的空间特征。设计C2f_MSC模块,提升多尺度特征提取能力,优化特征提取效率和网络表征能力,并采用轻量级CSFCN模块校准上下文和空间特征,实现对小目标特征信息的精细化处理。实验结果表明,改进算法在GLO-DET数据集上显著提高了目标缺陷的检测性能,尤其在检测低对比度缺陷和小目标缺陷中表现尤为突出。此外,在VisDrone2019Det和NEU-DET数据集上验证了改进算法的可泛化性。然而,尽管改进算法在检测精度上表现良好,但检测速度仍需提升,未来将重点研究轻量级网络,以实现工业产品的实时检测。
参考文献 引证文献
排序方式:
[1]
王伟,万晓刚.结合注意力机制和特征融合的小目标检测方法[J].西安工程大学学报,2022,36(6):115-123.
WANG W,WAN X G.Asmall target detection method combining attention mechanism and feature fusion[J].Journal of Xi'an Polytechnic University,2022, 36(6):115-123.(in Chinese)
[2]
WU X W,SAHOO D,HOI S C H.Recent advances in deep learning for object detection[J].Neurocomputing,2020,396:39-64.
[3]
ZHANG K H,SHEN H K.Solder joint defect detection in the connectors using improved faster-RCNN algorithm[J].Applied Sciences,2021,11(2):576.
[4]
YANG A M,JIANG T Y,HAN Y,et al.Research on application of on-line melting in-situ visual inspection of iron ore powder based on faster R-CNN[J].Alexandria Engineering Journal, 2022, 61 (11):8963-8971.
[5]
KUMAR A,MANIKANDAN R.Brain tumor detection using deep neural network-based classifier[C]//International Conference on Innovative Computing and Communications. Singapore: Springer, 2022:173-181.
[6]
LIU T,CHEN S Y.YOLOv4-DCN-based fabric defect detection algorithm[C]//2022 37th Youth Academic Annual Conference of Chinese Association of Automation.Beijing:IEEE,2023:710-715.
[7]
JIANG P Y,ERGU D,LIU F Y,et al.A review of yolo algorithm developments[J].Procedia Computer Science,2022,199:1066-1073.
[8]
江祥奎,杜遥遥,胡浩昌.一种改进YOLOv5s小目标无人机实时检测算法[J].西安邮电大学学报,2023, 28(3):88-96.
JIANG X K, DU Y Y, HU H C. An improved YOLOv5s real-time detection algorithm for small target UAVs[J].Journal of Xi'an University of Posts and Telecommunications,2023,28(3):88-96.(in Chinese)
[9]
DU B W,WAN F,LEI G B,et al.YOLO-MBBi:PCB surface defect detection method based on enhanced YOLOv5[J].Electronics,2023,12(13):2821.
[10]
YEUNG C C,LAM K M.Efficient fused-attention model for steel surface defect detection[J]. IEEE Transactions on Instrumentation and Measurement, 2022,71:2510011.
[11]
REN Y H,ZHANG T,MA R H,et al.Insulator defect detection method based on improved YOLOv7[C]//20239th International Conference on Mechanical and Electronics Engineering. Xi'an: IEEE, 2024:478-483.
[12]
WANG H B,WANG Y X.Improved glove defect detection algorithm based on YOLOv5 framework[C]//2022 IEEE 6th Advanced Information Technology,Electronic and Automation Control Conference.Beijing:IEEE,2022:1192-1197.
[13]
JIN H L,DU R Y,QIAO L Y,et al.CCA-YOLO:An improved glove defect detection algorithm based on YOLOv5[J].Applied Sciences,2023,13(18):10173.
[14]
CHU Z,WENG G,YU L.Real-time industrial surface defect detection based on lightweight convolutional neural networks[J].Artificial Intelligence and Machine Learning Review,2024,5(2):36-53.
[15]
江祥奎,杨刚,杜遥遥.基于轻量化YOLOv8n的动态视觉SLAM算法[J].西安邮电大学学报,2024,29 (3):75-82.
JIANG X K,YANG G,DU Y Y.Dynamic visual SLAM algorithm based on lightweight YOLOv8n[J]. Journal of Xi'an University of Posts and Telecommunications,2024,29(3):75-82.(in Chinese)
[16]
赵旭东,王馨,刘兴,.改进YOLOv5的拥挤场景目标检测算法[J].电子设计工程,2024,32(14):19-24.
ZHAO X D,WANG X,LIU X,et al.Improving YOLOv5's crowded scene object detection algorithm[J].Electronic Design Engineering,2024,32(14):19-24.(in Chinese)
[17]
HUANG Z L,ZHANG C X,GE L Y,et al.Joining spatial deformable convolution and a dense feature pyramid for surface defect detection[J].IEEE Transactions on Instrumentation and Measurement,2024,73:5012614.
[18]
SUN F X,HE N,WANG X,et al.YOLOv7-P:A lighter and more effective UAV aerial photography object detection algorithm[J].Signal,Image and Video Processing,2024,18(11):8327-8335.
[19]
WEI H,ZHAO L C,LI R P,et al.RFAConv-CBM-ViT:Enhanced vision transformer for metal surface defect detection[J].The Journal of Supercomputing, 2024,81(1):155.
[20]
YU W H,ZHOU P,YAN S C,et al.InceptionNeXt:When inception meets ConvNeXt[C]//2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Seattle:IEEE,2024:5672-5683.
[21]
SHI T J,GONG J N,HU J M,et al.Feature-enhanced CenterNet for small object detection in remote sensing images[J].Remote Sensing,2022,14 (21):5488.
[22]
YAN C Q,ZHANG H,LI X L,et al.R-SSD:Refined single shot multibox detector for pedestrian detection[J]. Applied Intelligence, 2022, 52 (9):10430-10447.
[23]
JIANG X B,GAO T H,ZHU Z C,et al.Real-time face mask detection method based on YOLOv3[J].Electronics,2021,10(7):837.
[24]
WANG C Y, BOCHKOVSKIY A, LIAO H M. YOLOv7:Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Vancouver:IEEE,2023:7464-7475.
[25]
WANG C Y, YEH I H, MARK LIAO H Y. YOLOv9:Learning what you want to learn using programmable gradient information[C]//Computer Vision-ECCV 2024.Cham:Springer,2025:1-21.
[26]
WANG A,CHEN H,LIU L,et al.YOLOv10:Real-time end-to-end object detection[J].Advances in Neural Information Processing Systems, 2025, 37:107984-108011.
2025年第30卷第6期
PDF下载
85
41
引用本文
BibTeX
文章信息
doi: 10.13682/j.issn.2095-6533.2025.06.014
  • 接收时间:2024-11-28
  • 首发时间:2026-04-16
  • 出版时间:2025-11-10
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-11-28
基金
作者信息
    1.西安邮电大学计算机学院,陕西西安 710121
    2.陕西省网络数据分析与智能处理重点实验室,陕西西安 710121
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/xayddxxb/CN/10.13682/j.issn.2095-6533.2025.06.014
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