Article(id=1156949466570904341, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156949362480861758, articleNumber=null, orderNo=null, doi=10.12404/j.issn.1671-1815.2402033, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1710950400000, receivedDateStr=2024-03-21, revisedDate=1732204800000, revisedDateStr=2024-11-22, acceptedDate=null, acceptedDateStr=null, onlineDate=1753767847911, onlineDateStr=2025-07-29, pubDate=1738944000000, pubDateStr=2025-02-08, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1753767847911, onlineIssueDateStr=2025-07-29, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1753767847911, creator=13701087609, updateTime=1753767847911, updator=13701087609, issue=Issue{id=1156949362480861758, tenantId=1146029695717560320, journalId=1146123166801305609, year='2025', volume='25', issue='4', pageStart='1312', pageEnd='1751', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=0, createTime=1753767823094, creator=13701087609, updateTime=1755171161273, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1162835389472424814, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156949362480861758, language=EN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1162835389472424815, tenantId=1146029695717560320, journalId=1146123166801305609, issueId=1156949362480861758, language=CN, specialIssueTitle=, coverIllustrator=, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=1701, endPage=1710, ext={EN=ArticleExt(id=1156949467858555690, articleId=1156949466570904341, tenantId=1146029695717560320, journalId=1146123166801305609, language=EN, title=Turning Traffic Flow Detection Method Based on Deep Learning, columnId=1156262728772735295, journalTitle=Science Technology and Engineering, columnName=Papers·Traffics and Transportations, runingTitle=null, highlight=null, articleAbstract=

In order to facilitate the counting of turning traffic flow and to enhance the detection speed and accuracy of turning traffic flow at intersections, a deep learning-based method was suggested for detecting, tracking, and counting turning traffic flow at urban crossings. Initially, the YOLOv5s, which was lightweight and efficient, was chosen as the target detection framework after conducting a comparative analysis. Unmanned aerial vehicle (UAV) aerial photography was utilized to record video footage of traffic movement at urban intersections, resulting in the development of a dataset of vehicle aerial photography photos. The pre-training weights and the most recent weight files were utilized to conduct training and testing on the self-constructed dataset. The model evaluation shows that the vehicle detection model using YOLOv5 exhibits great detection speed and accuracy. The model’s box_loss value declines rapidly and stabilizes at 0.038, while the mAP_0.5 value climbs swiftly and stays near 0.91.After that, the DeepSORT model was used as the backend multi-vehicle tracking technique, and a corner-to-centroid coordinate transformation was used to simplify the extraction of vehicle trajectories. The precision of the driving trajectory line was evaluated thereafter. To improve the robustness of trajectory points’ coordinate information, a corner-point-center-of-mass point coordinate transformation was suggested to tackle the issue of corner points in the detection frame. A sixth-degree polynomial was used to model the vehicle trajectory. Unsuitable trajectory lines were rotated and optimized to meet the function mapping requirements and ensure good fitting of all trajectories. Turning vehicles were detected and counted by using a predetermined threshold to determine the turning angle. Ultimately, to validate the performance of the proposed turning vehicle flow detection method, vehicle detection experiments were conducted at a city intersection as an illustration. The manual counting values were compared and analyzed against the detection results obtained using this method. The results show that the average detection accuracy for the four flow directions is 92.9%, with a maximum of 95.7%, meeting the standard detection requirements for turning vehicle flow in real intersection scenarios.

, correspAuthors=Yong LI, 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=Wei ZHANG, Yong LI, Tao LIU, Min-jie HA), CN=ArticleExt(id=1156949549362270900, articleId=1156949466570904341, tenantId=1146029695717560320, journalId=1146123166801305609, language=CN, title=基于深度学习的转弯车流量检测方法, columnId=1156262730664366426, journalTitle=科学技术与工程, columnName=论文·交通运输, runingTitle=null, highlight=null, articleAbstract=为方便统计转弯车流量,并提升交叉口转弯车流量的检测速度与精度,提出基于深度学习的方法对城市交叉口转弯车流量进行检测、跟踪和计数。首先,通过对比分析选用轻量高效的YOLOv5s作为目标检测框架,并采用无人机 (unmanned aerial vehicle,UAV)航拍方式获取城市交叉口交通流视频,自建车辆航拍图像数据集;利用预训练权重及最新权重文件完成自建数据集的训练与推理;模型评估表明,基于YOLOv5的车辆检测模型具有较高的检测速度与精度:其中模型的box_loss值迅速下降并稳定在0.038,mAP_0.5值迅速上升并保持在0.91附近;之后,对接DeepSORT模型作为后端多车辆跟踪算法,通过坐标转换以简化车辆轨迹提取,并对行驶轨迹线展开有效性判断;针对检测框角点跃变现象,提出角点-质心点坐标变换以强化轨迹点的坐标信息鲁棒性,且采用六次多项式拟合车辆轨迹线,将不满足函数映射要求的轨迹线进行旋转优化,以正常拟合全部轨迹;最后根据预设的转弯角度判定阈值,实现转弯车辆的检测与计数。为验证所提出的转弯车流量检测方法的有效性,以某一城市交叉口为例进行车辆检测实验,对比分析人工计数值和本方法检测结果。结果表明:4个流向平均检测精度为92.9%,最高可达95.7%,能够满足实际交叉口场景转弯车流量的常规检测要求。, correspAuthors=李永, authorNote=null, correspAuthorsNote=
*李永(2000—),男,汉族,河南南召人,硕士研究生。研究方向:道路交通安全改善及防治技术、智能交通。E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=QAu1NFkoqiO/NQ4E4Ka/Sw==, magXml=IgmDfEpSdBCkmEAICdLjKA==, pdfUrl=null, pdf=jkkWJcWeQL6Qusynm8/kzg==, pdfFileSize=11951400, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=DgkFJ144/SihZOgCsePMYA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=/L8EUXjPbZeYYVkwZ2n3lA==, mapNumber=null, authorCompany=null, fund=null, authors=

张韡(1973—),男,汉族,陕西榆林人,博士,副教授。研究方向:交通参与者行为干预、车辆建模与仿真、道路交通安全改善及防治技术、道路交通事故致因理论。E-mail:

, authorsList=张韡, 李永, 刘涛, 哈敏捷)}, authors=[Author(id=1225944418633560401, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=zhweijx@chd.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1225944419015242095, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, authorId=1225944418633560401, language=EN, stringName=Wei ZHANG, firstName=Wei, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 Key Laboratory of Transportation Industry of Automobile Transportation Safety and Security Technology, Chang’an University, Xi’an 710018, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1225944420651020674, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, authorId=1225944418633560401, language=CN, stringName=张韡, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018, bio={"content":"

张韡(1973—),男,汉族,陕西榆林人,博士,副教授。研究方向:交通参与者行为干预、车辆建模与仿真、道路交通安全改善及防治技术、道路交通事故致因理论。E-mail:

"}, bioImg=null, bioContent=

张韡(1973—),男,汉族,陕西榆林人,博士,副教授。研究方向:交通参与者行为干预、车辆建模与仿真、道路交通安全改善及防治技术、道路交通事故致因理论。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1225944418188964138, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, xref=1, ext=[AuthorCompanyExt(id=1225944418218324271, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418188964138, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Key Laboratory of Transportation Industry of Automobile Transportation Safety and Security Technology, Chang’an University, Xi’an 710018, China), AuthorCompanyExt(id=1225944418243490097, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418188964138, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018)])]), Author(id=1225944420856541592, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=3106757110@qq.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1225944421192085929, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, authorId=1225944420856541592, language=EN, stringName=Yong LI, firstName=Yong, middleName=null, lastName=LI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 Key Laboratory of Transportation Industry of Automobile Transportation Safety and Security Technology, Chang’an University, Xi’an 710018, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1225944421380829623, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, authorId=1225944420856541592, language=CN, stringName=李永, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1225944418188964138, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, xref=1, ext=[AuthorCompanyExt(id=1225944418218324271, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418188964138, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Key Laboratory of Transportation Industry of Automobile Transportation Safety and Security Technology, Chang’an University, Xi’an 710018, China), AuthorCompanyExt(id=1225944418243490097, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418188964138, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018)])]), Author(id=1225944421506658756, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1225944421636682198, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, authorId=1225944421506658756, language=EN, stringName=Tao LIU, firstName=Tao, middleName=null, lastName=LIU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 Key Laboratory of Transportation Industry of Automobile Transportation Safety and Security Technology, Chang’an University, Xi’an 710018, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1225944421741539812, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, authorId=1225944421506658756, language=CN, stringName=刘涛, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1225944418188964138, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, xref=1, ext=[AuthorCompanyExt(id=1225944418218324271, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418188964138, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Key Laboratory of Transportation Industry of Automobile Transportation Safety and Security Technology, Chang’an University, Xi’an 710018, China), AuthorCompanyExt(id=1225944418243490097, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418188964138, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018)])]), Author(id=1225944421871563249, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, 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=1225944422001586687, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, authorId=1225944421871563249, language=EN, stringName=Min-jie HA, firstName=Min-jie, middleName=null, lastName=HA, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 Xi’an BYD Auto Co., Ltd., Xi’an 710018, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1225944422127415824, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, authorId=1225944421871563249, language=CN, stringName=哈敏捷, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2 西安比亚迪汽车有限公司, 西安 710018, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1225944418386096440, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, xref=2, ext=[AuthorCompanyExt(id=1225944418474176832, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418386096440, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Xi’an BYD Auto Co., Ltd., Xi’an 710018, China), AuthorCompanyExt(id=1225944418486759745, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418386096440, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 西安比亚迪汽车有限公司, 西安 710018)])])], keywords=[Keyword(id=1225944422462960174, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, orderNo=1, keyword=traffic information engineering), Keyword(id=1225944422580400701, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, orderNo=2, keyword=traffic flow detection), Keyword(id=1225944422685258314, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, orderNo=3, keyword=deep learning), Keyword(id=1225944422836253270, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, orderNo=4, keyword=turning vehicle), Keyword(id=1225944422970471010, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, orderNo=5, keyword=object detection), Keyword(id=1225944423142437502, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, orderNo=6, keyword=vehicle tracking), Keyword(id=1225944423226323592, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, orderNo=1, keyword=交通信息工程), Keyword(id=1225944423394095771, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, orderNo=2, keyword=交通流检测), Keyword(id=1225944424937599671, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, orderNo=3, keyword=深度学习), Keyword(id=1225944425059234503, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, orderNo=4, keyword=转弯车辆), Keyword(id=1225944425260561125, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, orderNo=5, keyword=目标检测), Keyword(id=1225944425390584569, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, orderNo=6, keyword=车辆跟踪)], refs=[Reference(id=1225944431266803933, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=152, pageEnd=162, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=Bui K H N, Yi H, Jung H E J, journalName=12th Asian Conference on Intelligent Information and Database Systems, refType=null, unstructuredReference=Bui K H N, Yi H, Jung H E J, et al. Video based traffic flow analysis for turning volume estimation at signalized intersections[C]// 12th Asian Conference on Intelligent Information and Database Systems. Seoul: IEEE, 2020: 152-162., articleTitle=Video based traffic flow analysis for turning volume estimation at signalized intersections, refAbstract=null), Reference(id=1225944431421993191, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=28, pageStart=12159, pageEnd=12167, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=蒲玲玲, 杨柳, journalName=科学技术与工程, refType=null, unstructuredReference=蒲玲玲, 杨柳. 改进YOLOv5的多车辆目标实时检测及跟踪算法[J]. 科学技术与工程, 2023, 23(28): 12159-12167., articleTitle=改进YOLOv5的多车辆目标实时检测及跟踪算法, refAbstract=null), Reference(id=1225944431556210929, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=23, issue=28, pageStart=12159, pageEnd=12167, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=Pu Lingling, Yang Liu, journalName=Science Techno-logy and Engineering, refType=null, unstructuredReference=Pu Lingling, Yang Liu. Improved real-time detection and tracking algorithm for multi vehicle targets in YOLOv5[J]. Science Techno-logy and Engineering, 2023, 23(28): 12159-12167., articleTitle=Improved real-time detection and tracking algorithm for multi vehicle targets in YOLOv5, refAbstract=null), Reference(id=1225944431723983108, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=1991, volume=null, issue=null, pageStart=43, pageEnd=54, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=Turk M A, Pent A P, journalName=Intelligent Robots and Computer Vision IX, refType=null, unstructuredReference=Turk M A, Pent A P. Recognition in face space[C]// Intelligent Robots and Computer Vision IX: Algorithms and Techniques. International Society for Optics and Photonics. New York: IEEE, 1991: 43-54., articleTitle=Recognition in face space, refAbstract=null), Reference(id=1225944431879172370, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2011, volume=15, issue=null, pageStart=315, pageEnd=323, url=null, language=null, rfNumber=[4], rfOrder=4, authorNames=Glorot X, Bordes A, Bengio Y, journalName=Journal of Machine Learning Research, refType=null, unstructuredReference=Glorot X, Bordes A, Bengio Y. Deep sparse rectifier neural networks[J]. Journal of Machine Learning Research, 2011, 15: 315-323., articleTitle=Deep sparse rectifier neural networks, refAbstract=null), Reference(id=1225944432017584408, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[5], rfOrder=5, authorNames=Howard A G, Zhu M, Chen B, journalName=arXiv Preprint, refType=null, unstructuredReference=Howard A G, Zhu M, Chen B, et al. MobileNets: efficient convolutional neural networks for mobile vision applications[J]. arXiv Preprint, 2017: DOI:10.48550/arXiv.1704.04861.., articleTitle=MobileNets: efficient convolutional neural networks for mobile vision applications, refAbstract=null), Reference(id=1225944432164385060, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2019, volume=572, issue=null, pageStart=106, pageEnd=111, url=null, language=null, rfNumber=[6], rfOrder=6, authorNames=Pei J, Deng L, Song S, journalName=Nature, refType=null, unstructuredReference=Pei J, Deng L, Song S, et al. Towards artificial general intelligence with hybrid Tianjic chip architecture[J]. Nature, 2019, 572: 106-111., articleTitle=Towards artificial general intelligence with hybrid Tianjic chip architecture, refAbstract=null), Reference(id=1225944432340545842, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2021, volume=21, issue=8, pageStart=3192, pageEnd=3198, url=null, language=null, rfNumber=[7], rfOrder=7, authorNames=袁小平, 马绪起, 刘赛, journalName=科学技术与工程, refType=null, unstructuredReference=袁小平, 马绪起, 刘赛. 改进YOLOv3的行人车辆目标检测算法[J]. 科学技术与工程, 2021, 21(8): 3192-3198., articleTitle=改进YOLOv3的行人车辆目标检测算法, refAbstract=null), Reference(id=1225944433707888960, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2021, volume=21, issue=8, pageStart=3192, pageEnd=3198, url=null, language=null, rfNumber=[7], rfOrder=8, authorNames=Yuan Xiaoping, Ma Xuqi, Liu Sai, journalName=Science Technology and Engineering, refType=null, unstructuredReference=Yuan Xiaoping, Ma Xuqi, Liu Sai. An improved algorithm of pedestrian and vehicle detection based on YOLOv3[J]. Science Technology and Engineering, 2021, 21(8): 3192-3198., articleTitle=An improved algorithm of pedestrian and vehicle detection based on YOLOv3, refAbstract=null), Reference(id=1225944433821135178, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2024, volume=54, issue=8, pageStart=2338, pageEnd=2347, url=null, language=null, rfNumber=[8], rfOrder=9, authorNames=才华, 寇婷婷, 杨依宁, journalName=吉林大学学报(工学版), refType=null, unstructuredReference=才华, 寇婷婷, 杨依宁, 等. 基于轨迹优化的三维车辆多目标跟踪[J]. 吉林大学学报(工学版), 2024, 54(8): 2338-2347., articleTitle=基于轨迹优化的三维车辆多目标跟踪, refAbstract=null), Reference(id=1225944433976324430, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2024, volume=54, issue=8, pageStart=2338, pageEnd=2347, url=null, language=null, rfNumber=[8], rfOrder=10, authorNames=Cai Hua, Kou Tingting, Yang Yining, journalName=Journal of Jinlin University(Engineering and Technology Edition), refType=null, unstructuredReference=Cai Hua, Kou Tingting, Yang Yining, et al. Three-dimensional vehicle multi-target tracking based on trajectory optimization[J]. Journal of Jinlin University(Engineering and Technology Edition), 2024, 54(8): 2338-2347., articleTitle=Three-dimensional vehicle multi-target tracking based on trajectory optimization, refAbstract=null), Reference(id=1225944434194428253, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[9], rfOrder=11, authorNames=Bewley A, Ge Z, Ott L, journalName=IEEE International Conference on Image Processing (ICIP), refType=null, unstructuredReference=Bewley A, Ge Z, Ott L, et al. Simple online andrealtime tracking[C]// IEEE International Conference on Image Processing (ICIP). New York: IEEE, 2016: DOI:10.1109/ICIP. 2016.7533003., articleTitle=Simple online andrealtime tracking, refAbstract=null), Reference(id=1225944434311868771, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[10], rfOrder=12, authorNames=Wojke N, Bewley A, Paulus D, journalName=null, refType=null, unstructuredReference=Wojke N, Bewley A, Paulus D. Simple online and realtime tracking with a deep association metric[C]// 2017 IEEE International Conference on Image Processing (ICIP). New York: IEEE, 2017: DOI:10.1109/icip. 2017.8296962., articleTitle=Simple online and realtime tracking with a deep association metric, refAbstract=null), Reference(id=1225944434517389681, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2021, volume=55, issue=6, pageStart=1056, pageEnd=1064, url=null, language=null, rfNumber=[11], rfOrder=13, authorNames=金立生, 华强, 郭柏苍, journalName=浙江大学学报(工学版), refType=null, unstructuredReference=金立生, 华强, 郭柏苍, 等. 基于优化DeepSort的前方车辆多目标跟踪[J]. 浙江大学学报(工学版), 2021, 55(6): 1056-1064., articleTitle=基于优化DeepSort的前方车辆多目标跟踪, refAbstract=null), Reference(id=1225944434651607420, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2021, volume=55, issue=6, pageStart=1056, pageEnd=1064, url=null, language=null, rfNumber=[11], rfOrder=14, authorNames=Jin Lisheng, Hua Qiang, Guo Baicang, journalName=Journal of Zhejiang University (Engineering Science), refType=null, unstructuredReference=Jin Lisheng, Hua Qiang, Guo Baicang, et al. Multi-target tracking of vehicles based on optimized DeepSort[J]. Journal of Zhejiang University (Engineering Science), 2021, 55(6): 1056-1064., articleTitle=Multi-target tracking of vehicles based on optimized DeepSort, refAbstract=null), Reference(id=1225944434773242246, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=36, issue=9, pageStart=288, pageEnd=301, url=null, language=null, rfNumber=[12], rfOrder=15, authorNames=程鑫, 周经美, 刘霈源, journalName=中国公路学报, refType=null, unstructuredReference=程鑫, 周经美, 刘霈源, 等. 基于任务联合的三维车辆检测与跟踪集成算法[J]. 中国公路学报, 2023, 36(9): 288-301., articleTitle=基于任务联合的三维车辆检测与跟踪集成算法, refAbstract=null), Reference(id=1225944434903265681, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=36, issue=9, pageStart=288, pageEnd=301, url=null, language=null, rfNumber=[12], rfOrder=16, authorNames=Cheng Xin, Zhou Jingmei, Liu Peiyuan, journalName=China Journal of Highway and Transport, refType=null, unstructuredReference=Cheng Xin, Zhou Jingmei, Liu Peiyuan, et al. 3D vehicle detection and tracking integration algorithm based on task collaboration[J]. China Journal of Highway and Transport, 2023, 36(9): 288-301., articleTitle=3D vehicle detection and tracking integration algorithm based on task collaboration, refAbstract=null), Reference(id=1225944434991346075, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2015, volume=37, issue=9, pageStart=1904, pageEnd=1916, url=null, language=null, rfNumber=[13], rfOrder=17, authorNames=He K, Zhang X, Ren S, journalName=IEEE Transactions on Pattern Analysis & Machine Intelligence, refType=null, unstructuredReference=He K, Zhang X, Ren S, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2015, 37(9): 1904-1916., articleTitle=Spatial pyramid pooling in deep convolutional networks for visual recognition, refAbstract=null), Reference(id=1225944435096203685, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=43, issue=3, pageStart=95, pageEnd=104, url=null, language=null, rfNumber=[14], rfOrder=18, authorNames=柴艳娜, 宋焕生, 朱婧, journalName=长安大学学报(自然科学版), refType=null, unstructuredReference=柴艳娜, 宋焕生, 朱婧. 航拍图像车辆检测的数据增强[J]. 长安大学学报(自然科学版), 2023, 43(3): 95-104., articleTitle=航拍图像车辆检测的数据增强, refAbstract=null), Reference(id=1225944435234615726, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=43, issue=3, pageStart=95, pageEnd=104, url=null, language=null, rfNumber=[14], rfOrder=19, authorNames=Chai Yanna, Song Huansheng, Zhu Jing, journalName=Journal of Chang’an University(Natural Science Edition), refType=null, unstructuredReference=Chai Yanna, Song Huansheng, Zhu Jing. Data augmentation on vehicle detection in aerial-images[J]. Journal of Chang’an University(Natural Science Edition), 2023, 43(3): 95-104., articleTitle=Data augmentation on vehicle detection in aerial-images, refAbstract=null), Reference(id=1225944435352056241, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2022, volume=113, issue=null, pageStart=104914, pageEnd=null, url=null, language=null, rfNumber=[15], rfOrder=20, authorNames=Dong X D, Yan S, Duan C Q, journalName=Engineering Applications of Artificial Intelligence, refType=null, unstructuredReference=Dong X D, Yan S, Duan C Q. A lightweight vehicles detection network model based on YOLOv5[J]. Engineering Applications of Artificial Intelligence, 2022, 113: 104914., articleTitle=A lightweight vehicles detection network model based on YOLOv5, refAbstract=null), Reference(id=1225944435482079673, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2021, volume=21, issue=4, pageStart=48, pageEnd=71, url=null, language=null, rfNumber=[16], rfOrder=21, authorNames=马永杰, 程时升, 马芸婷, journalName=交通运输工程学报, refType=null, unstructuredReference=马永杰, 程时升, 马芸婷, 等. 卷积神经网络及其在智能交通系统中的应用综述[J]. 交通运输工程学报, 2021, 21(4): 48-71., articleTitle=卷积神经网络及其在智能交通系统中的应用综述, refAbstract=null), Reference(id=1225944435603714499, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2021, volume=21, issue=4, pageStart=48, pageEnd=71, url=null, language=null, rfNumber=[16], rfOrder=22, authorNames=Ma Yongjie, Cheng Shisheng, Ma Yunting, journalName=Journal of Traffic and Transportation Engineering, refType=null, unstructuredReference=Ma Yongjie, Cheng Shisheng, Ma Yunting, et al. Review of convolutional neural network and its application in intelligent transportation system[J]. Journal of Traffic and Transportation Engineering, 2021, 21(4): 48-71., articleTitle=Review of convolutional neural network and its application in intelligent transportation system, refAbstract=null), Reference(id=1225944435758903754, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2011, volume=34, issue=7, pageStart=1409, pageEnd=1422, url=null, language=null, rfNumber=[17], rfOrder=23, authorNames=Kalal Z, Mikolajczyk K, Matas J, journalName=IEEE Transactions on Software Engineering, refType=null, unstructuredReference=Kalal Z, Mikolajczyk K, Matas J. Tracking-learning-detection[J]. IEEE Transactions on Software Engineering, 2011, 34(7): 1409-1422., articleTitle=Tracking-learning-detection, refAbstract=null), Reference(id=1225944435893121491, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=45, issue=4, pageStart=674, pageEnd=682, url=null, language=null, rfNumber=[18], rfOrder=24, authorNames=贾志, 李茂军, 李婉婷, journalName=计算机工程与科学, refType=null, unstructuredReference=贾志, 李茂军, 李婉婷. 基于改进YOLOv5+DeepSort算法模型的交叉路口车辆实时检测[J]. 计算机工程与科学, 2023, 45(4): 674-682., articleTitle=基于改进YOLOv5+DeepSort算法模型的交叉路口车辆实时检测, refAbstract=null), Reference(id=1225944436056699353, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=45, issue=4, pageStart=674, pageEnd=682, url=null, language=null, rfNumber=[18], rfOrder=25, authorNames=Jia Zhi, Li Maojun, Li Wanting, journalName=Computer Engineering & Science, refType=null, unstructuredReference=Jia Zhi, Li Maojun, Li Wanting. Real-time vehicle detection at intersections based on improved YOLOv5+DeepSort algorithm mode[J]. Computer Engineering & Science, 2023, 45(4): 674-682., articleTitle=Real-time vehicle detection at intersections based on improved YOLOv5+DeepSort algorithm mode, refAbstract=null), Reference(id=1225944436182528482, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=26, authorNames=哈敏捷, journalName=基于深度学习算法的转弯车流量检测研究, refType=null, unstructuredReference=哈敏捷. 基于深度学习算法的转弯车流量检测研究[D]. 西安: 长安大学, 2022., articleTitle=null, refAbstract=null), Reference(id=1225944436299969000, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=27, authorNames=Ha Minjie, journalName=Research on turning traffic flow detection based on deep learning algorithm, refType=null, unstructuredReference=Ha Minjie. Research on turning traffic flow detection based on deep learning algorithm[D]. Xi’an: Chang’an University, 2022., articleTitle=null, refAbstract=null), Reference(id=1225944436434186734, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2000, volume=null, issue=4, pageStart=465, pageEnd=472, url=null, language=null, rfNumber=[20], rfOrder=28, authorNames=王云诚, 唐焕文, journalName=计算数学, refType=null, unstructuredReference=王云诚, 唐焕文. 单峰函数最优化问题的进化策略[J]. 计算数学, 2000(4): 465-472., articleTitle=单峰函数最优化问题的进化策略, refAbstract=null), Reference(id=1225944436572598775, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2000, volume=null, issue=4, pageStart=465, pageEnd=472, url=null, language=null, rfNumber=[20], rfOrder=29, authorNames=Wang Yuncheng, Tang Huanwen, journalName=Mathematica Numerica Sinica, refType=null, unstructuredReference=Wang Yuncheng, Tang Huanwen. An evoluti-onary strategy for minimzing unimodal functi-ons[J]. Mathematica Numerica Sinica, 2000(4): 465-472., articleTitle=An evoluti-onary strategy for minimzing unimodal functi-ons, refAbstract=null), Reference(id=1225944436706816511, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=142, pageEnd=147, url=null, language=null, rfNumber=[21], rfOrder=30, authorNames=Mirthubashini J, Santhi V, journalName=null, refType=null, unstructuredReference=Mirthubashini J, Santhi V. 6th International Conference on Advanced Computing andCommunication Systems (ICACCS). Video basedvehicle counting using deep learning algorithms[C]// New York: IEEE, 2020: 142-147., articleTitle=6th International Conference on Advanced Computing andCommunication Systems (ICACCS). Video basedvehicle counting using deep learning algorithms, refAbstract=null), Reference(id=1225944436820062725, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[22], rfOrder=31, authorNames=Meng C S, Yi X W, Su L, journalName=null, refType=null, unstructuredReference=Meng C S, Yi X W, Su L, et al. City-wide traffic volume inference with loop detectordata and taxi trajectories[C]// 25th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS). New York: ACM, 2017: DOI:10.1145/3139958.3139984., articleTitle=City-wide traffic volume inference with loop detectordata and taxi trajectories, refAbstract=null), Reference(id=1225944438204183051, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=51, issue=10, pageStart=68, pageEnd=88, url=null, language=null, rfNumber=[23], rfOrder=32, authorNames=徐志航, 么新鹏, 徐志刚, journalName=华南理工大学学报(自然科学版), refType=null, unstructuredReference=徐志航, 么新鹏, 徐志刚, 等. 道路交通检测器及其优化布设方法研究综述[J]. 华南理工大学学报(自然科学版), 2023, 51(10): 68-88., articleTitle=道路交通检测器及其优化布设方法研究综述, refAbstract=null), Reference(id=1225944438325817877, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2023, volume=51, issue=10, pageStart=68, pageEnd=88, url=null, language=null, rfNumber=[23], rfOrder=33, authorNames=Xu Zhihang, Yao Xinpeng, Xu Zhigang, journalName=Journal of South China University of Technology (Natural Science Edition), refType=null, unstructuredReference=Xu Zhihang, Yao Xinpeng, Xu Zhigang, et al. Review of research on road traffic detectorsand its optimized deployment methods[J]. Journal of South China University of Technology (Natural Science Edition), 2023, 51(10): 68-88., articleTitle=Review of research on road traffic detectorsand its optimized deployment methods, refAbstract=null), Reference(id=1225944438439064092, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2009, volume=26, issue=9, pageStart=235, pageEnd=238, url=null, language=null, rfNumber=[24], rfOrder=34, authorNames=熊文华, 徐建闽, 林思, journalName=计算机仿真, refType=null, unstructuredReference=熊文华, 徐建闽, 林思. 基于BP网络的浮动车与线圈检测数据融合模型[J]. 计算机仿真, 2009, 26 (9): 235-238., articleTitle=基于BP网络的浮动车与线圈检测数据融合模型, refAbstract=null), Reference(id=1225944438581670437, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, doi=null, pmid=null, pmcid=null, year=2009, volume=26, issue=9, pageStart=235, pageEnd=238, url=null, language=null, rfNumber=[24], rfOrder=35, authorNames=Xiong Wenhua, Xu Jianmin, Lin Si, journalName=Computer Simulation, refType=null, unstructuredReference=Xiong Wenhua, Xu Jianmin, Lin Si. A fusionmodelfor floating vehicle and loop detector data based on BP network[J]. Computer Simulation, 2009, 26(9): 235-238., articleTitle=A fusionmodelfor floating vehicle and loop detector data based on BP network, refAbstract=null)], funds=[Fund(id=1225944430981591244, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, awardId=51978075, language=CN, fundingSource=国家自然科学基金面上项目(51978075), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1225944418188964138, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, xref=1, ext=[AuthorCompanyExt(id=1225944418218324271, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418188964138, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 Key Laboratory of Transportation Industry of Automobile Transportation Safety and Security Technology, Chang’an University, Xi’an 710018, China), AuthorCompanyExt(id=1225944418243490097, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418188964138, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018)]), AuthorCompany(id=1225944418386096440, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, xref=2, ext=[AuthorCompanyExt(id=1225944418474176832, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418386096440, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 Xi’an BYD Auto Co., Ltd., Xi’an 710018, China), AuthorCompanyExt(id=1225944418486759745, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, companyId=1225944418386096440, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2 西安比亚迪汽车有限公司, 西安 710018)])], figs=[ArticleFig(id=1225944425696768802, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.1, caption=Network structure of YOLOv5s, figureFileSmall=bOHNka9/JlJ83Q4IQRk/zw==, figureFileBig=yALEuwzSBWsM+ywO7vyUPA==, tableContent=null), ArticleFig(id=1225944425826792241, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图1, caption=YOLOv5s网络结构图, figureFileSmall=bOHNka9/JlJ83Q4IQRk/zw==, figureFileBig=yALEuwzSBWsM+ywO7vyUPA==, tableContent=null), ArticleFig(id=1225944426028118860, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.2, caption=Process for creating self-build vehicle dataset, figureFileSmall=wl6E7aZTaF63khVG5Od5oQ==, figureFileBig=haWJGMJ42Fe9jDQB61xB1Q==, tableContent=null), ArticleFig(id=1225944426116199256, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图2, caption=自制车辆数据集流程, figureFileSmall=wl6E7aZTaF63khVG5Od5oQ==, figureFileBig=haWJGMJ42Fe9jDQB61xB1Q==, tableContent=null), ArticleFig(id=1225944426279777132, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.3, caption=Comparison of vehicle detection effect, figureFileSmall=ztK0adImKMyF+eY2lLYCUQ==, figureFileBig=HfVpG1VxNV6I/A4bi7VDNg==, tableContent=null), ArticleFig(id=1225944426397217658, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图3, caption=车辆检测效果对比, figureFileSmall=ztK0adImKMyF+eY2lLYCUQ==, figureFileBig=HfVpG1VxNV6I/A4bi7VDNg==, tableContent=null), ArticleFig(id=1225944426548212620, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.4, caption=Performance verification ofYOLOv5s, figureFileSmall=F/0IT+DO1K2xcifsonXeXw==, figureFileBig=cH9uR9s8nvN8gUNgvouXyw==, tableContent=null), ArticleFig(id=1225944426690818968, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图4, caption=模型的性能验证

mAP_0.5表示IoU阈值为0.5时的平均精确度。

, figureFileSmall=F/0IT+DO1K2xcifsonXeXw==, figureFileBig=cH9uR9s8nvN8gUNgvouXyw==, tableContent=null), ArticleFig(id=1225944426971837367, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.5, caption=Flowchart of multi-vehicle tracking, figureFileSmall=gGWUJP+ItNyW77z+cwZKHQ==, figureFileBig=1R+WUHNEh6kwM8z9JOEW2g==, tableContent=null), ArticleFig(id=1225944427202524112, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图5, caption=多车辆跟踪流程图, figureFileSmall=gGWUJP+ItNyW77z+cwZKHQ==, figureFileBig=1R+WUHNEh6kwM8z9JOEW2g==, tableContent=null), ArticleFig(id=1225944427303187421, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.6, caption=Detection of the turning process for vehicle ID-10, figureFileSmall=R+5zqyh936wxkaUGclRP+Q==, figureFileBig=VmX//dAsVy8GJXFbAZyzVw==, tableContent=null), ArticleFig(id=1225944427412239344, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图6, caption=车辆ID-10检测转弯过程, figureFileSmall=R+5zqyh936wxkaUGclRP+Q==, figureFileBig=VmX//dAsVy8GJXFbAZyzVw==, tableContent=null), ArticleFig(id=1225944427538067456, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.7, caption=Schematic diagram of rotated coordinates, figureFileSmall=+419eyBLMI1VyoFiR9mRsw==, figureFileBig=BC1f0IUqc0aahV6LT8li+g==, tableContent=null), ArticleFig(id=1225944427672285202, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图7, caption=旋转坐标示意图, figureFileSmall=+419eyBLMI1VyoFiR9mRsw==, figureFileBig=BC1f0IUqc0aahV6LT8li+g==, tableContent=null), ArticleFig(id=1225944427798114337, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.8, caption=Comparison of fitting effect, figureFileSmall=MoadUbpWrh1fsqP/1Mrj3Q==, figureFileBig=ZzKNS8zf0xH7xPUsicy3sA==, tableContent=null), ArticleFig(id=1225944427894583342, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图8, caption=拟合效果对比, figureFileSmall=MoadUbpWrh1fsqP/1Mrj3Q==, figureFileBig=ZzKNS8zf0xH7xPUsicy3sA==, tableContent=null), ArticleFig(id=1225944429270315069, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.9, caption=Schematic diagram of judgment conditions for turning, figureFileSmall=jT5tH8OFpcvq4MoMrBEtdg==, figureFileBig=VSQ9SP79xMfDFnjfJgjtdw==, tableContent=null), ArticleFig(id=1225944429471641680, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图9, caption=转弯判定条件示意图, figureFileSmall=jT5tH8OFpcvq4MoMrBEtdg==, figureFileBig=VSQ9SP79xMfDFnjfJgjtdw==, tableContent=null), ArticleFig(id=1225944429651996763, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Fig.10, caption=Determination of the intersection to which the vehicle belongs, figureFileSmall=NQ3FOEDZgkC3Edib3vXjOA==, figureFileBig=MiAPzaVe1P69oq5y9ruIJw==, tableContent=null), ArticleFig(id=1225944429870100586, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=图10, caption=车辆所属路口判定, figureFileSmall=NQ3FOEDZgkC3Edib3vXjOA==, figureFileBig=MiAPzaVe1P69oq5y9ruIJw==, tableContent=null), ArticleFig(id=1225944430050455673, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Table 1, caption=

Goodness of fit value for partial vehicle track lines

, figureFileSmall=null, figureFileBig=null, tableContent=
车辆轨迹线ID 拟合优度值
ID-14 0.973
ID-37 0.930
ID-50 0.926
ID-91 0.945
ID-108 0.918
), ArticleFig(id=1225944430365028494, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=表1, caption=

部分车辆轨迹线拟合优度值

, figureFileSmall=null, figureFileBig=null, tableContent=
车辆轨迹线ID 拟合优度值
ID-14 0.973
ID-37 0.930
ID-50 0.926
ID-91 0.945
ID-108 0.918
), ArticleFig(id=1225944430562160799, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=EN, label=Table 2, caption=

Test results for turning vehicles

, figureFileSmall=null, figureFileBig=null, tableContent=
路口 转向 检测车流
量/辆
实际车流
量/辆
误检数
/辆
漏检数
/辆
准确
率/%
路口A 左转 41 44 0 3 93.2
右转 57 62 1 4 91.9
路口B 左转 25 27 0 2 92.6
右转 37 41 1 3 90.2
路口C 左转 22 23 0 1 95.7
右转 35 38 1 2 92.1
路口D 左转 17 18 0 1 94.4
右转 41 44 0 3 93.2
), ArticleFig(id=1225944430721544365, tenantId=1146029695717560320, journalId=1146123166801305609, articleId=1156949466570904341, language=CN, label=表2, caption=

转弯车辆检测结果

, figureFileSmall=null, figureFileBig=null, tableContent=
路口 转向 检测车流
量/辆
实际车流
量/辆
误检数
/辆
漏检数
/辆
准确
率/%
路口A 左转 41 44 0 3 93.2
右转 57 62 1 4 91.9
路口B 左转 25 27 0 2 92.6
右转 37 41 1 3 90.2
路口C 左转 22 23 0 1 95.7
右转 35 38 1 2 92.1
路口D 左转 17 18 0 1 94.4
右转 41 44 0 3 93.2
)], 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.1671-1815.2402033, detailUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/10.12404/j.issn.1671-1815.2402033, pdfUrlCn=https://castjournals.cast.org.cn/joweb/kxjsygc/CN/PDF/10.12404/j.issn.1671-1815.2402033, pdfUrlEn=https://castjournals.cast.org.cn/joweb/kxjsygc/EN/PDF/10.12404/j.issn.1671-1815.2402033, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
基于深度学习的转弯车流量检测方法
收藏切换
PDF下载
张韡 1 , 李永 1, * , 刘涛 1 , 哈敏捷 2
科学技术与工程 | 论文·交通运输 2025,25(4): 1701-1710
收起
收藏切换
科学技术与工程 | 论文·交通运输 2025, 25(4): 1701-1710
基于深度学习的转弯车流量检测方法
全屏
张韡1 , 李永1, * , 刘涛1, 哈敏捷2
作者信息
  • 1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018
  • 2 西安比亚迪汽车有限公司, 西安 710018
  • 张韡(1973—),男,汉族,陕西榆林人,博士,副教授。研究方向:交通参与者行为干预、车辆建模与仿真、道路交通安全改善及防治技术、道路交通事故致因理论。E-mail:

通讯作者:

*李永(2000—),男,汉族,河南南召人,硕士研究生。研究方向:道路交通安全改善及防治技术、智能交通。E-mail:
Turning Traffic Flow Detection Method Based on Deep Learning
Wei ZHANG1 , Yong LI1, * , Tao LIU1, Min-jie HA2
Affiliations
  • 1 Key Laboratory of Transportation Industry of Automobile Transportation Safety and Security Technology, Chang’an University, Xi’an 710018, China
  • 2 Xi’an BYD Auto Co., Ltd., Xi’an 710018, China
出版时间: 2025-02-08 doi: 10.12404/j.issn.1671-1815.2402033
文章导航
收藏切换
为方便统计转弯车流量,并提升交叉口转弯车流量的检测速度与精度,提出基于深度学习的方法对城市交叉口转弯车流量进行检测、跟踪和计数。首先,通过对比分析选用轻量高效的YOLOv5s作为目标检测框架,并采用无人机 (unmanned aerial vehicle,UAV)航拍方式获取城市交叉口交通流视频,自建车辆航拍图像数据集;利用预训练权重及最新权重文件完成自建数据集的训练与推理;模型评估表明,基于YOLOv5的车辆检测模型具有较高的检测速度与精度:其中模型的box_loss值迅速下降并稳定在0.038,mAP_0.5值迅速上升并保持在0.91附近;之后,对接DeepSORT模型作为后端多车辆跟踪算法,通过坐标转换以简化车辆轨迹提取,并对行驶轨迹线展开有效性判断;针对检测框角点跃变现象,提出角点-质心点坐标变换以强化轨迹点的坐标信息鲁棒性,且采用六次多项式拟合车辆轨迹线,将不满足函数映射要求的轨迹线进行旋转优化,以正常拟合全部轨迹;最后根据预设的转弯角度判定阈值,实现转弯车辆的检测与计数。为验证所提出的转弯车流量检测方法的有效性,以某一城市交叉口为例进行车辆检测实验,对比分析人工计数值和本方法检测结果。结果表明:4个流向平均检测精度为92.9%,最高可达95.7%,能够满足实际交叉口场景转弯车流量的常规检测要求。
交通信息工程  /  交通流检测  /  深度学习  /  转弯车辆  /  目标检测  /  车辆跟踪

In order to facilitate the counting of turning traffic flow and to enhance the detection speed and accuracy of turning traffic flow at intersections, a deep learning-based method was suggested for detecting, tracking, and counting turning traffic flow at urban crossings. Initially, the YOLOv5s, which was lightweight and efficient, was chosen as the target detection framework after conducting a comparative analysis. Unmanned aerial vehicle (UAV) aerial photography was utilized to record video footage of traffic movement at urban intersections, resulting in the development of a dataset of vehicle aerial photography photos. The pre-training weights and the most recent weight files were utilized to conduct training and testing on the self-constructed dataset. The model evaluation shows that the vehicle detection model using YOLOv5 exhibits great detection speed and accuracy. The model’s box_loss value declines rapidly and stabilizes at 0.038, while the mAP_0.5 value climbs swiftly and stays near 0.91.After that, the DeepSORT model was used as the backend multi-vehicle tracking technique, and a corner-to-centroid coordinate transformation was used to simplify the extraction of vehicle trajectories. The precision of the driving trajectory line was evaluated thereafter. To improve the robustness of trajectory points’ coordinate information, a corner-point-center-of-mass point coordinate transformation was suggested to tackle the issue of corner points in the detection frame. A sixth-degree polynomial was used to model the vehicle trajectory. Unsuitable trajectory lines were rotated and optimized to meet the function mapping requirements and ensure good fitting of all trajectories. Turning vehicles were detected and counted by using a predetermined threshold to determine the turning angle. Ultimately, to validate the performance of the proposed turning vehicle flow detection method, vehicle detection experiments were conducted at a city intersection as an illustration. The manual counting values were compared and analyzed against the detection results obtained using this method. The results show that the average detection accuracy for the four flow directions is 92.9%, with a maximum of 95.7%, meeting the standard detection requirements for turning vehicle flow in real intersection scenarios.

traffic information engineering  /  traffic flow detection  /  deep learning  /  turning vehicle  /  object detection  /  vehicle tracking
张韡, 李永, 刘涛, 哈敏捷. 基于深度学习的转弯车流量检测方法. 科学技术与工程, 2025 , 25 (4) : 1701 -1710 . DOI: 10.12404/j.issn.1671-1815.2402033
Wei ZHANG, Yong LI, Tao LIU, Min-jie HA. Turning Traffic Flow Detection Method Based on Deep Learning[J]. Science Technology and Engineering, 2025 , 25 (4) : 1701 -1710 . DOI: 10.12404/j.issn.1671-1815.2402033
基于视频图像的交通量检测[1]作为缓解交通堵塞问题的重要支持手段,对提升交通安全和强化交通组织有着重要的实际意义和社会应用价值,在交通运输领域一直备受关注。但传统检测方法费时费力,随着深度学习算法的日渐成熟,越来越多的学者将基于深度学习的目标检测和跟踪算法引入交通流检测领域[2],以实现运动车辆的识别和跟踪。
Truk等[3]将Eigenface方法应用于人脸分类,该算法属于早期的目标检测的基本构想。随后,基于手工特征的传统目标检测算法得到迅速发展,Glorot等[4]将ReLU函数引入深度学习,并证实了该激活函数可显著提升神经网络的非线性拟合能力,在特定条件下可以完全克服sigmoid和tanh等激活函数的饱和问题,为之后的深度学习模型训练奠定了坚实基础。基于深度学习的目标检测逐渐成为主流,研究人员提出了多种检测算法。Howard等[5]提出了MobileNet,其作为一个轻量级的卷积神经网络自带两个可调的超参数,可针对特定的任务和实际应用场景需要调整模型的延迟和精度。清华大学类脑计算研究中心于2019年自行研发出中国首款类脑芯片,该芯片能够支持脉冲神经网络和人工神经网络同时运行[6],使得中国在视觉目标检测、目标跟踪等众多应用开发项目上具备了较大优势。袁小平等[7]通过将YOLOv3的原生残差模块替换为ResneXt,并在卷积网络Darknet-53中引入密集连接,实现了多尺度特征的高效提取和充分利用,改进后的DX-LOVO算法在中小目标平均检测准确率及精度方面取得了有效提升。
目标检测技术经过多年发展,在多数复杂环境中均可获得出色的检测效果。但现阶段的跟踪算法仍存在一定缺陷,尤其是应用于运动车辆识别领域的多目标跟踪(multi-object-tracking,MOT)算法,该算法在复杂应用场景下存在检测准确率低,实时性较差等问题[8],导致多目标跟踪研究在交通量检测领域进展缓慢。随着计算机视觉和深度学习的融合发展,越来越多的研究人员针对多目标跟踪算法的优化和应用开展研究。Bewley等[9]提出了一种简单实用、实时性较强的SORT(simple online and realtime tracking)算法,该方法基于Faster RCNN(faster region convolutional neural networks)检测框架,在目标跟踪时未使用目标的外观特征,而是结合卡尔曼滤波和匈牙利算法,以分别实现目标运动估计和数据关联,由于SORT算法仅通过IoU(intersection over union)进行匹配,虽然其检测精度和速度均得到了较大提升,但该算法仍无法较好解决物体遮挡问题。Wojke等[10]在SORT算法的基础上提出DeepSORT算法,该算法将面积匹配修改为特征匹配,采用一个小型CNN(convolutional neural networks)提取并保存被检测目标的外观特征,这种改进有效降低了物体被遮挡后的ID switch现象。金立生等[11]以YOLOv3作为前端目标检测器,引入中心损失函数以改进深度余弦度量,实验结果表明,优化后的DeepSORT算法在拥挤交通和遮挡环境下具有良好的跟踪效果,与原生DeepSORT算法相比,多目标跟踪准确率提升1%,ID switch现象减少4%。此类基于检测的跟踪算法遵循“先检测后跟踪”设计思想[12],前端目标检测的准确性将会对后端的目标跟踪产生决定性作用。
基于视频图像的交通量检测按照流向可分为直线车流量检测和转弯车辆检测,目前前人研究多是针对直线车辆检测统计,即基于目标检测算法完成车辆识别并采用撞虚拟线圈方式进行后端计数工作。但关于交叉路口转弯车辆检测统计算法屈指可数,缺少一套完善有效的检测体系。
目前,关于交叉口转弯车流量的研究较少,前人研究主要集中在基于简单交通场景下的车辆目标单一检测,且检测车辆行驶轨迹的研究多处于识别直行车辆阶段。而交叉口转弯车流量作为城市交叉口流量调查环节的重要指标,对交叉口的交通组织和信号配时优化具有重要意义,却鲜有研究关注。鉴于此,提出一种基于深度学习的方法,以YOLOv5s作为前端车辆目标检测器,结合后端DeepSORT跟踪算法,对转弯车辆的行驶轨迹进行分析处理,并根据预设转弯角度阈值和计数器,以期实现转弯车辆检测与流量统计。
YOLOv5的网络结构作为经典的one-stage结构,主要由输入端、Backbone、Neck及Prediction组成。图像在输入端经过Mosaic数据增强[13-14]、自适应锚框计算和图片放缩后完成预处理;Backbone中的Focus则是YOLOv5独有的结构,通过一系列切片操作完成特征提取,以提升网络速度;Neck作为连接Backbone和输出端的中间模块,进行深层次的特征融合和上采样操作,但区别于YOLOv4只在主干网络中使用CSP结构,YOLOv5s设计一种CSP2_X结构应用于Neck,强化了网络的特征融合能力。其网络结构如图1所示。
由于YOLOv5采用更加灵活的网络结构,可以根据不同实际需求对主干网络进行调整,使其在保证较高的检测速度及检测精度的同时,检测画面更加清晰,且优化了重叠边框等问题[15],YOLOv5s其网络深度和特征图宽度又是同系列中最小,检测速度显著高于其他版本的同时亦可拥有较好的检测精度。因此采用YOLOv5s作为车辆目标检测框架。
随着深度学习在交通领域的不断融合发展,与交通场景相关的开源图像数据集越来越多,如BDD100K自动驾驶数据集、Waymo开放数据集、MIT DriveSeg Dataset数据集等,为智能交通发展提供了帮助[16]。但对于交叉口转弯车辆检测而言,目前此类开源数据集的图像数据拍摄视角都是基于车载摄像头或路侧监控,不适合本研究使用,因此需要自行制作交叉口行车环境下转弯车辆图像数据集,作为模型训练样本。
将采集到的多段城市交叉口航拍视频以视频分帧的方式输出为多张1 920像素×1 080像素的图片,以此作为原始图像数据集。考虑本研究的重点是转弯车辆的识别与跟踪,故不着重考虑无车辆或车辆被遮挡等特殊情况处理,因此将原始图像集中的上述无效数据剔除,最终获得8 651张有效图像,将其编码后借助LabelImg为图像数据标注车辆标签,得到最终自制数据集,具体制作流程如图2所示。
将自建数据集以8∶2的比例划分为训练集和测试集,且仅需识别车辆目标,属于单分类任务,类别数设置为1,训练轮次“epochs”默认300,单次输入图片“batch_size”视配置修改为4。
利用YOLOv5s自带的预训练权重完成模型训练后,使用上述已收敛模型产生的最新权重文件对测试集进行推理测试,模型的预训练结果及推理测试效果如图3所示。
从检出效果看,图3(a)中虽有少量车辆未被检出,但相较于原生YOLOv5s模型在小目标检测方面的平均精度有较大改善,对大部分车辆目标能够实现较高精度的正确识别,且被识别车辆的平均置信度约为0.8,模型推理速度也显著提升。
为进一步验证YOLOv5s检测模型在真实交叉口场景的车辆检测性能。对自建数据集中全部图像进行300轮次训练,并考虑到转弯车辆检测是二分类问题,且数据集具有不平衡特性,故采用平均精确度(mean average precision,mAP)作为模型评估指标,模型在检测过程中的损失变化和平均精度变化如图4所示。
图4中可以看出,模型在训练初期的边界框损失快速下降而平均精度则快速上升,这是由于在训练过程中使用预训练权重作为迁移学习起点,加速了YOLOv5s的训练速度;而在该范围内两指标均出现震荡现象,分析可能是自建数据集样本量较少受训练数据噪声影响较大所致。随着训练轮次的增加,模型的训练损失和验证损失均变化趋势几乎一致,均表现为逐渐下降并趋于稳定;验证损失在既定训练轮次内并未出现增大现象,故模型不存在过拟合现象;训练损失最终保持在0.038,同时平均精度随之逐渐增加并稳定在0.91附近,两个指标均表明该模型具有较好的车辆检测性能,可以满足后续转弯车辆跟踪研究。
采用TBD(tracking-by-detecton)[17]策略以实现多车辆跟踪,即使用YOLOv5算法作为车辆检测器,在此基础上对接DeepSORT完成车辆预测与跟踪,并基于跟踪车辆的轨迹特征信息,结合提出的车辆转弯事件检测模型,以完成实际交叉口场景下的转弯车流量检测。
对比SORT算法,DeepSORT在匹配机制中引入了级联匹配算法,因此能更好处理目标被长时间遮挡及ID切换问题[18],故采用DeepSORT算法作为多车辆跟踪模型,并基于匀速模型和线性观测模型下的卡尔曼滤波器完成运动车辆轨迹预测及更新。
将前端目标检测器检测到的车辆坐标信息输入至DeepSORT检测器,完成坐标格式转换后基于卡尔曼滤波实现逐帧预测,若预测结果是确认态则利用马氏距离和余弦距离使检测框和预测框产生数据关联,如式(1)、式(2)所示。
${\stackrel{·}{d}}_{(i,j)}=({d}_{j}-{y}_{i}{)}^{T}{S}_{i}^{-1}({d}_{j}-{y}_{i})$
式(1)中:${\stackrel{·}{d}}_{(i,j)}$为第j个检测框与第i个预测框间的运动匹配度;dj为第j个检测框的状态;yi为当前时刻预测框状态;${S}_{i}^{-1}$为当前观测空间的协方差矩阵。
${\ddot{d}}_{(i,j)}=min\{1-{{r}^{T}}_{j}{r}_{k}^{i}|{r}_{k}^{i}\in {R}_{i}\}$
式(2)中:rj${r}_{k}^{i}$分别为相对应的检测框和预测框的表面特征描述因子;Ri为车辆轨迹的描述因子集合。
为了更好满足实际场景下的检测需求,考虑将车辆的运动特征度量和外观特征度量相结合,融合指标的计算公式为
${d}_{(i,j)}=\lambda {\stackrel{·}{d}}_{(i,j)}+(1-\lambda){\ddot{d}}_{(i,j)}$
式(3)中:$\lambda $为权重参数。
后续在关联状态下,使用检测框替代上一时刻的跟踪框进行更新。循环执行上述预测、跟踪和更新流程,即可完成不发生ID switch的车辆匹配。
但考虑到实际情况中无法保证车辆检测的召回率为100%,因此利用DeepSORT对匹配失败的检测框和预测框进行IoU匹配;若匹配成功,则继续车辆跟踪、预测。对于再次匹配失败的预测框,如果当前处于为非确认状态,则将其删除;若为确认状态,则通过检查寿命时间来决定是否继续匹配。基于DeepSORT实现多车辆跟踪的具体流程如图5所示。
在交叉口交通流量较大时,对多车辆连续帧之间的关系建模存在较大难度,而且目前针对此交通环境下转弯车辆的检测存在诸多弊端。因此,基于DeepSORT以视频分帧读取的方式追踪和定位车辆ID,之后在图像坐标系描绘目标点集,使用OpenCV绘制车辆轨迹并进行曲线拟合,即可提取出车辆的行驶轨迹线,以供后续检测转弯车辆。
在提取车辆行驶轨迹过程中,针对车辆检测失败导致轨迹线断开,或视频边缘区域车辆停留时间短等原因造成行驶轨迹线过短等实际问题,使用欧式距离对车辆轨迹线长度进行有效性判定,计算公式为
$\rho =\sqrt{({x}_{2}-{x}_{1}{)}^{2}+({y}_{2}-{y}_{1}{)}^{2}}$
式(4)中:(x1,y1)为轨迹线起点坐标;(x2,y2)为终点坐标;$\rho $为两点间的欧氏距离。
经过大量实验分析,判别阈值$\rho $取200个单位长度时筛选效果较好[19]。超过该阈值,则认定提取的轨迹线能够代表车辆的实际运动轨迹;否则认为该曲线不完整,予以舍弃。
经过上述长度有效性筛选后,对车辆轨迹线的处理步骤如下。
步骤1 建立坐标系:以图像左上角为原点建立笛卡尔坐标系,逐次对全部轨迹线进行坐标化处理,以便提取各轨迹线的位置信息。
步骤2 获取车辆坐标:车辆在转弯过程中后端跟踪器实时获取前端YOLOv5生成的检测框状态(ymin,xmin,w,ℎ),并依此持续进行全部车辆ID的定位采集。其中,xmin、ymin分别为检测框左上顶点的横、纵坐标;w、ℎ分别为检测框的宽、高。标记ID号为l的车辆坐标信息为$({x}^{l},{y}^{l},w,ℎ),{x}^{l}、{y}^{l}$分别为ID号为l的车辆横、纵位置。图6展示了ID-10号的车辆位置信息检测过程。
当车辆转弯时,由于检测框会发生方向性变形和长宽比例畸变,其角点坐标会随之发生跃变,从而导致提取的车辆轨迹线会出现较多偏差点。为了应对这一异常情况,采用质心点代替角点,即在原始角点坐标的基础上引入偏置量,以增强轨迹点的坐标信息鲁棒性,转换公式为
$\left\{\begin{array}{l}x={x}_{i}+\frac{{w}_{i}}{2}\\ y={y}_{i}+\frac{{ℎ}_{i}}{2}\end{array}\right.$
式(5)中:xi,yi为检测框角点坐标;x,y为质心点坐标;wii分别为检测框的宽和高。
步骤3 绘制车辆轨迹线:依据车辆ID逐帧提取车辆轨迹点坐标${V}_{n}^{l}=\{{V}_{1},{V}_{2},\dots,{V}_{n}\}$,并将轨迹线绘制在图像上,为后续轨迹拟合提供数据基础。
采用六次多项式对车辆轨迹线进行拟合,并利用优度指标评估拟合效果,详细流程见步骤5。拟合多项式表示为
$\left\{\begin{array}{l}\varphi (x,\omega)={X}_{1\times 7}W\\ {X}_{1\times 7}=({x}^{6},{x}^{5},{x}^{4},{x}^{3},{x}^{2},x,1)\\ {{W}^{T}}_{1\times 7}=({\omega }_{6},{\omega }_{5},{\omega }_{4},{\omega }_{3},{\omega }_{2},{\omega }_{1},{\omega }_{0})\end{array}\right.$
式(6)中:$\varphi $为轨迹拟合多项式;$x、\omega $为多项式函数的自变量与多项式系数;X1×7、W1×7为多项式的自变量阶次向量与系数向量。
步骤4 拟合多项式的坐标变换:由于轨迹线直接绘制在图像上,车辆实际的转弯轨迹拟合曲线不满足“多对一”的函数映射条件,但考虑到多数车辆轨迹线都形如单峰曲线[20],故对轨迹线进行坐标变换使其满足函数映射要求。
对于无法直接拟合的车辆轨迹,将采集到的车辆轨迹点集逐点按照图7所示方式,进行顺时针旋转操作;通过定义旋转角度步长,逐步迭代发现旋转角度为10°时效果最佳。之后基于步骤5对轨迹进行拟合优度评价,循环上述评价-旋转-再评价操作,直至拟合出的函数图像完全贴合车辆轨迹线结束该操作。
图7展示了图像坐标系中一点P关于原点进行旋转的过程,旋转角为$\theta $,若实际情况需要逆时针旋转,将$\theta $改为$-\theta $代入式(7)即可。经过上述循环操作即可得到成功的拟合函数。
$\left\{\begin{array}{l}cos\left(\theta \right)x-sin\left(\theta \right)y\\ cos\left(\theta \right)y+sin\left(\theta \right)x\end{array}\right.$
为了验证采用六次多项式拟合车辆转弯轨迹的可行性及所提出的坐标变换修正轨迹拟合方法的有效性,分别基于四次、五次、六次和七次多项式对旋转函数图像进行曲线拟合,拟合效果分别如图8所示。其中,四次多项式的拟合效果最差,无法实现车辆轨迹的正常拟合;五次多项式在轨迹线中间部分拟合效果较好,但在轨迹线两端的拟合效果出现偏离异常趋势,分析判断是车辆在进入图像瞬间,检测框尺寸瞬间变大,从而表现出明显抖动,故会导致检出的车辆坐标与真实轨迹线产生部分拟合异常。相比之下,基于六次和七次多项式拟合图像几乎完美贴合转弯轨迹线,但考虑到多项式阶次过高会增加过拟合风险,故采用六次多项式进行车辆转弯轨迹拟合。
步骤5 轨迹线拟合优度评价。
拟合优度是回归分析中用以评价拟合曲线对观测值拟合程度的常用指标,其统计量确定系数R2的计算公式为
$\left\{\begin{array}{l}TSS=\sum _{i=1}^{m}({y}_{i}{-\stackrel{-}{y})}^{2}\\ RSS=\sum _{i=1}^{m}({\widehat{y}}_{i}-{y}_{i}{)}^{2}\\ {R}^{2}=1-\frac{RSS}{TSS}\end{array}\right.$
式(8)中:m为拟合轨迹点数量;yi为原始轨迹点纵坐标;$\stackrel{-}{y}$为所有轨迹点纵坐标平均值;${\widehat{y}}_{i}$为多项式拟合所得轨迹点纵坐标;TSS为总平方和;RSS为残差平方和;R2为确定系数。
R2的计算结果处于[0,1]范围内,数值越大代表拟合效果越好,根据拟合优度检验可以对车辆轨迹线的拟合效果进行量化评价,为修正轨迹拟合提供支持。表1展示了随机抽取部分车辆ID的轨迹拟合优度评价结果。
表1可知,随机选取的5个车辆ID的轨迹拟合度均超过0.91,由此可见基于TBD策略的交叉口转弯车辆检测方法可以对转弯车辆行驶轨迹线实现较高的拟合性能,可视作后续转弯车辆行驶轨迹的计算依据。
经过上述流程即可得到多组完整的交叉口车辆转弯轨迹线,将目标轨迹线的两端点分别记为I0、I1,其对应拟合多项式函数的起点、终点分别记为$\varphi \left({I}_{0}\right)、\varphi \left({I}_{1}\right)。$由于六次多项式拟合如前所述会充分贴合车辆轨迹线,可能也存在过拟合现象,若直接对I0求梯度会导致与车辆实际行驶方向偏差较大,不利于后续车辆转弯角度判定。故采用一阶向前差分代替梯度求解车辆的行驶方向变化角度。以车辆轨迹线起点I0为例,其一阶向前差分计算公式为
$d({I}_{0},{I}_{0}+\Delta I)=\frac{\varphi \left({I}_{0}\right)-\varphi ({I}_{0}+\Delta I)}{\Delta I}$
式(9)中:$d({I}_{0},{I}_{0}+\Delta I)$为起点I0的一阶向前差分值;$\Delta I$为步长,取5。
车辆完成整个转弯动作两端点的的角度改变量$\Delta \theta $可表示为
$\left\{\begin{array}{l}\Delta \theta ={\theta }_{0}-{\theta }_{1}\\ {\theta }_{0}=arctan\left[d\right({I}_{0},{I}_{0}+\Delta I\left)\right]\\ {\theta }_{1}=arctan\left[d\right({I}_{1},{I}_{1}+\Delta I\left)\right]\end{array}\right.$
式(10)中:${\theta }_{0}、{\theta }_{1}$分别为车辆换道轨迹起始时刻及终止时刻的行驶方向变化角。
以车辆转弯动作的角度改变量变化范围作为车辆转弯事件的判定准则,如图9所示。并按照多次实验结果将这一阈值定为60°。若$\Delta \theta 60°$,判断目标车辆为转弯车辆,令相应计数器+1,且当$\Delta \theta 0$时,车辆为右转;当$\Delta \theta 0$时,车辆为左转。计数器取值如式(11)所示。
${N}_{v}=\left\{\begin{array}{ll}1,& \Delta \theta 0\\ 0,& \mathrm{其}\mathrm{他}\end{array}\right.$
式(11)中:Nv为车辆转弯事件计数结果。
最后基于虚拟线圈法[21]实现转弯车辆所属路口的判定问题。如图10所示,在交叉口道路图像的各进口道停止线处构建一个30像素宽的矩形框区域,并将矩形框按照顺时针方向依次记为A、B、C、D;其中,图10左侧矩形框为B。当目标框的预设撞线点经过矩形框时即采集目标车辆ID。
为验证基于YOLOv5+DeepSORT算法的交叉口转弯车流量检测模型的有效性和准确度,选取西安市某一信控交叉口进行实际测试。利用大疆“御”Mavic 2无人机获取该交叉口航拍交通流视频,相机参数:视频分辨率为1 920×1 080,帧率为30 fps;拍摄时间:周内白天,天气晴朗,单次拍摄时长为15 min,且模型测试时以人工计数为准。
本实验的模型训练平台硬件:台式计算机,操作系统为Windows11(64Bit),处理器为Intel i7@3.4 GHz 4Cores,显卡为NVIDIA GeForce RTX3070 8 GB。运行环境及相关工具:Cuda10.2,Cudnn7605,Pytorch1.2,Python3.9,Anaconda3以及视觉工具Opencv2。
基于所建立模型对交叉口场景下转弯车流量进行统计测试,检测结果如表2所示。实际转弯车流量为人工计数获得,检测准确率以该数据为参照。从表2中数据可以看出,本文模型对交叉口4个流向转弯车流量的总体检测准确率均超过90%,平均准确率达到92.9%,其中,对路口C左转车流检测准确率最高达到了95.7%。虽然路口A、B的右转方向存在误检漏检情况,但经过分析,推断是因为数据采集时相机垂直拍摄角度存在一定倾斜,引起远离相机的上述路口待测车辆目标尺寸变小、特征提取更加困难,从而导致目标检测算法的精确率和召回率下降。但模型的平均检测准确率接近93%,可以满足实际场景中的检测精度需求。
相比于所提出的转弯车流量检测模型,适用于交叉口场景下的传统固定式交通流检测器检测精度虽大多可超过95%,但线圈检测器、地磁检测器等通常需要铺设在道路下方,会对道路造成一定程度破坏,其模块组成复杂,安装成本及维护成本极高[22],且相关研究表明传统固定式检测器的合理布设是获取高精度数据的关键因素,简单的数量堆叠既不经济也不合理[23]。除此之外,该类检测器受限于自身检测原理,当交通密度较大时,检测精度会显著下降[24]。此外,超声波检测器或红外检测器虽然不需要破坏路面即可完成布设,维护成本较低,但其检测精度仅能达到约90%,且容易受极端天气等干扰[23]。相比之下,利用无人机获取交叉口航拍视频并基于深度学习检测转弯车流量具有移动性强、操作简单、成本低等显著优势,可以在城市道路交通量检测方面发挥巨大作用。
(1)自制交叉口场景航拍车辆图像数据集,并采用Mosaic数据增强方式强化车辆检测网络的特征提取能力,在前端框架推理测试环节,基于YOLOv5s的车辆检测表现优异,平均精度超过90%,为后端车辆跟踪提供了精度保障。
(2)车辆跟踪阶段,采用TBD策略实现车辆跟踪,并利用六次多项式拟合车辆轨迹,引入坐标变换解决了转弯轨迹曲线不满足函数映射要求的问题,拟合优度评价结果显示,车辆转弯轨迹线的平均拟合度最高可以达到0.973。
(3)为验证提出的转弯车辆检测统计模型,基于某一城市交叉口进行实例分析,检测结果表明:交叉口4个进口道的转弯车流量平均检测准确率可达92.9%,可实现实际场景下转弯车流量的常规检测,为交叉口的精细化渠化及信号配时提供技术支持。
  • 国家自然科学基金面上项目(51978075)
参考文献 引证文献
排序方式:
[1]
Bui K H N, Yi H, Jung H E J, et al. Video based traffic flow analysis for turning volume estimation at signalized intersections[C]// 12th Asian Conference on Intelligent Information and Database Systems. Seoul: IEEE, 2020: 152-162.
[2]
蒲玲玲, 杨柳. 改进YOLOv5的多车辆目标实时检测及跟踪算法[J]. 科学技术与工程, 2023, 23(28): 12159-12167.
Pu Lingling, Yang Liu. Improved real-time detection and tracking algorithm for multi vehicle targets in YOLOv5[J]. Science Techno-logy and Engineering, 2023, 23(28): 12159-12167.
[3]
Turk M A, Pent A P. Recognition in face space[C]// Intelligent Robots and Computer Vision IX: Algorithms and Techniques. International Society for Optics and Photonics. New York: IEEE, 1991: 43-54.
[4]
Glorot X, Bordes A, Bengio Y. Deep sparse rectifier neural networks[J]. Journal of Machine Learning Research, 2011, 15: 315-323.
[5]
Howard A G, Zhu M, Chen B, et al. MobileNets: efficient convolutional neural networks for mobile vision applications[J]. arXiv Preprint, 2017: DOI:10.48550/arXiv.1704.04861..
[6]
Pei J, Deng L, Song S, et al. Towards artificial general intelligence with hybrid Tianjic chip architecture[J]. Nature, 2019, 572: 106-111.
[7]
袁小平, 马绪起, 刘赛. 改进YOLOv3的行人车辆目标检测算法[J]. 科学技术与工程, 2021, 21(8): 3192-3198.
Yuan Xiaoping, Ma Xuqi, Liu Sai. An improved algorithm of pedestrian and vehicle detection based on YOLOv3[J]. Science Technology and Engineering, 2021, 21(8): 3192-3198.
[8]
才华, 寇婷婷, 杨依宁, 等. 基于轨迹优化的三维车辆多目标跟踪[J]. 吉林大学学报(工学版), 2024, 54(8): 2338-2347.
Cai Hua, Kou Tingting, Yang Yining, et al. Three-dimensional vehicle multi-target tracking based on trajectory optimization[J]. Journal of Jinlin University(Engineering and Technology Edition), 2024, 54(8): 2338-2347.
[9]
Bewley A, Ge Z, Ott L, et al. Simple online andrealtime tracking[C]// IEEE International Conference on Image Processing (ICIP). New York: IEEE, 2016: DOI:10.1109/ICIP. 2016.7533003.
[10]
Wojke N, Bewley A, Paulus D. Simple online and realtime tracking with a deep association metric[C]// 2017 IEEE International Conference on Image Processing (ICIP). New York: IEEE, 2017: DOI:10.1109/icip. 2017.8296962.
[11]
金立生, 华强, 郭柏苍, 等. 基于优化DeepSort的前方车辆多目标跟踪[J]. 浙江大学学报(工学版), 2021, 55(6): 1056-1064.
Jin Lisheng, Hua Qiang, Guo Baicang, et al. Multi-target tracking of vehicles based on optimized DeepSort[J]. Journal of Zhejiang University (Engineering Science), 2021, 55(6): 1056-1064.
[12]
程鑫, 周经美, 刘霈源, 等. 基于任务联合的三维车辆检测与跟踪集成算法[J]. 中国公路学报, 2023, 36(9): 288-301.
Cheng Xin, Zhou Jingmei, Liu Peiyuan, et al. 3D vehicle detection and tracking integration algorithm based on task collaboration[J]. China Journal of Highway and Transport, 2023, 36(9): 288-301.
[13]
He K, Zhang X, Ren S, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2015, 37(9): 1904-1916.
[14]
柴艳娜, 宋焕生, 朱婧. 航拍图像车辆检测的数据增强[J]. 长安大学学报(自然科学版), 2023, 43(3): 95-104.
Chai Yanna, Song Huansheng, Zhu Jing. Data augmentation on vehicle detection in aerial-images[J]. Journal of Chang’an University(Natural Science Edition), 2023, 43(3): 95-104.
[15]
Dong X D, Yan S, Duan C Q. A lightweight vehicles detection network model based on YOLOv5[J]. Engineering Applications of Artificial Intelligence, 2022, 113: 104914.
[16]
马永杰, 程时升, 马芸婷, 等. 卷积神经网络及其在智能交通系统中的应用综述[J]. 交通运输工程学报, 2021, 21(4): 48-71.
Ma Yongjie, Cheng Shisheng, Ma Yunting, et al. Review of convolutional neural network and its application in intelligent transportation system[J]. Journal of Traffic and Transportation Engineering, 2021, 21(4): 48-71.
[17]
Kalal Z, Mikolajczyk K, Matas J. Tracking-learning-detection[J]. IEEE Transactions on Software Engineering, 2011, 34(7): 1409-1422.
[18]
贾志, 李茂军, 李婉婷. 基于改进YOLOv5+DeepSort算法模型的交叉路口车辆实时检测[J]. 计算机工程与科学, 2023, 45(4): 674-682.
Jia Zhi, Li Maojun, Li Wanting. Real-time vehicle detection at intersections based on improved YOLOv5+DeepSort algorithm mode[J]. Computer Engineering & Science, 2023, 45(4): 674-682.
[19]
哈敏捷. 基于深度学习算法的转弯车流量检测研究[D]. 西安: 长安大学, 2022.
Ha Minjie. Research on turning traffic flow detection based on deep learning algorithm[D]. Xi’an: Chang’an University, 2022.
[20]
王云诚, 唐焕文. 单峰函数最优化问题的进化策略[J]. 计算数学, 2000(4): 465-472.
Wang Yuncheng, Tang Huanwen. An evoluti-onary strategy for minimzing unimodal functi-ons[J]. Mathematica Numerica Sinica, 2000(4): 465-472.
[21]
Mirthubashini J, Santhi V. 6th International Conference on Advanced Computing andCommunication Systems (ICACCS). Video basedvehicle counting using deep learning algorithms[C]// New York: IEEE, 2020: 142-147.
[22]
Meng C S, Yi X W, Su L, et al. City-wide traffic volume inference with loop detectordata and taxi trajectories[C]// 25th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS). New York: ACM, 2017: DOI:10.1145/3139958.3139984.
[23]
徐志航, 么新鹏, 徐志刚, 等. 道路交通检测器及其优化布设方法研究综述[J]. 华南理工大学学报(自然科学版), 2023, 51(10): 68-88.
Xu Zhihang, Yao Xinpeng, Xu Zhigang, et al. Review of research on road traffic detectorsand its optimized deployment methods[J]. Journal of South China University of Technology (Natural Science Edition), 2023, 51(10): 68-88.
[24]
熊文华, 徐建闽, 林思. 基于BP网络的浮动车与线圈检测数据融合模型[J]. 计算机仿真, 2009, 26 (9): 235-238.
Xiong Wenhua, Xu Jianmin, Lin Si. A fusionmodelfor floating vehicle and loop detector data based on BP network[J]. Computer Simulation, 2009, 26(9): 235-238.
2025年第25卷第4期
PDF下载
290
107
引用本文
BibTeX
文章信息
doi: 10.12404/j.issn.1671-1815.2402033
  • 接收时间:2024-03-21
  • 首发时间:2025-07-29
  • 出版时间:2025-02-08
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-03-21
  • 修回日期:2024-11-22
基金
国家自然科学基金面上项目(51978075)
作者信息
    1 长安大学汽车运输安全保障技术交通行业重点实验室, 西安 710018
    2 西安比亚迪汽车有限公司, 西安 710018

通讯作者:

*李永(2000—),男,汉族,河南南召人,硕士研究生。研究方向:道路交通安全改善及防治技术、智能交通。E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/kxjsygc/CN/10.12404/j.issn.1671-1815.2402033
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