Article(id=1329420287659061894, tenantId=1146029695717560320, journalId=1324759914146529478, issueId=1329420206000157303, articleNumber=null, orderNo=null, doi=10.7544/issn1000-1239.202550696, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1758384000000, receivedDateStr=2025-09-21, revisedDate=1776960000000, revisedDateStr=2026-04-24, acceptedDate=1777478400000, acceptedDateStr=2026-04-30, onlineDate=1794888096011, onlineDateStr=2026-11-17, pubDate=1785513600000, pubDateStr=2026-08-01, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1794888096011, onlineIssueDateStr=2026-11-17, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1794888096011, creator=13701087609, updateTime=1794888096011, updator=13701087609, issue=Issue{id=1329420206000157303, tenantId=1146029695717560320, journalId=1324759914146529478, year='2026', volume='63', issue='8', pageStart='1905', pageEnd='2181', issueExtLink='null', onlineDate='null', pubDate='1785513600000', pubDateStr='2026-08-01', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1794888076542, creator='13701087609', updateTime=1794888076542, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext=null, issueFiles=null, downloadFileDto=null}, startPage=2076, endPage=2094, ext={EN=ArticleExt(id=1329420287877165703, articleId=1329420287659061894, tenantId=1146029695717560320, journalId=1324759914146529478, language=EN, title=3Deus: Object Recognition Method Based on Point Cloud Detection in Complex Scenario, columnId=1325489891789910943, journalTitle=Journal of Computer Research and Development, columnName=Artificial Intelligence, runingTitle=null, highlight=null, articleAbstract=

With the development of artificial intelligence technology, vision-based object recognition, as an important component of the Internet of things perception layer, can be applied in various cargo sorting scenarios when combined with sorting robots. However, in complex environments where objects are disorganized, relying solely on object recognition algorithms may not be sufficient to distinguish target objects from interfering ones, as the difference between them may be the spatial distance. This can negatively impact the robot’s normal grasping actions. To address this issue, we propose a lightweight object recognition method for complex environments—3Deus. This method combines deep learning with 3D point cloud data processing. It uses a depth camera to synchronously capture both RGB images and 3D point cloud data in the recognition scene. The method eliminates the plane where interfering objects are located using the 3D point cloud, and distinguishes the plane containing the target object from the stacked interfering objects with an improved RANSAC algorithm. It also employs a plane expansion mechanism to ensure the integrity of the plane point cloud where the target object is located. Finally, the object recognition algorithm is used to obtain the recognition bounding box of the target object and construct its mapping relationship with the specific point cloud cluster. We deploy 3Deus on edge computing devices and validate its effectiveness and robustness by constructing different experimental scenarios. Experimental results show that compared with the traditional random sampling consensus algorithm, 3Deus can effectively distinguish the target objects from interference objects, achieving an accuracy of over 95%. Moreover, by utilizing the plane expansion mechanism, the recognition accuracy of the target object is improved by 30.48%.

, authors=Zhichao Shao1, Jiefan Qiu1, Guanglin Dai1, 3, *, Qiaohong Li1, Hanmo Chen1, Kejiang Xiao2, authorsList=Zhichao Shao, Jiefan Qiu, Guanglin Dai, Qiaohong Li, Hanmo Chen, Kejiang Xiao, authorCompany=null, correspAuthors=Guanglin Dai, authorNote=

born in 2001. Master candidate. Student member of CCF. His main research interests include object detection and artificial intelligence

born in 1984. PhD, associate professor. Member of CCF. His main research interests include embedded operating systems, Internet of things, and artificial intelligence

born in 1979. PhD, professor. Member of CCF. His main research interests include intelligent computing, medical image processing, and big data analysis

born in 2004. Bachelor. His main research interests include artificial intelligence and object detection

born in 2001. Bachelor. His main research interests include machine vision and artificial intelligence

born in 1984. PhD, professor. His main research interests include multi-source information fusion, artificial intelligence, intelligent perception interconnection, and Internet of things systems

, correspAuthorsNote=null, copyrightStatement=Copyright © 2026 Journal of Computer Research and Development. All rights reserved., 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, fund=null), CN=ArticleExt(id=1329420294445445825, articleId=1329420287659061894, tenantId=1146029695717560320, journalId=1324759914146529478, language=CN, title=3Deus:基于点云探测的复杂场景物品识别方法, columnId=1325489892104483745, journalTitle=计算机研究与发展, columnName=人工智能, runingTitle=null, highlight=null, articleAbstract=

随着人工智能技术的发展,基于视觉的物品识别作为构建物联网感知层的重要组成部分,与分拣机器人相结合后可应用在各类货物分拣的具体场景中。然而,在混乱摆放物品的复杂场景中,由于目标物品与干扰物品可能存在空间距离差,仅依赖物品识别算法并不能很好地区分目标物品和干扰物品,势必对机器人的正常抓取动作产生负面影响。为了解决该问题,提出了一种轻量化的复杂环境下的物品识别方法——3Deus。该方法将深度学习和3D点云数据处理相结合,利用3D点云对干扰物品所处的平面做整体消除,并基于改进的随机采样一致性算法区分目标物品所处平面与堆叠干扰物;运用平面扩张机制,保证目标物品所处平面点云的完整性;最后,借助物品识别算法获取目标的识别框并构建其与具体点云簇的映射关系。将3Deus部署在边缘计算设备上,并通过构建不同的实验场景验证了3Deus的有效性和鲁棒性。实验结果表明,与传统的随机采样一致性算法相比,3Deus能够有效地对目标物品和干扰物品进行区分,其准确率达到了95%以上,同时3Deus利用平面扩张机制将目标物品的识别准确率提高了30.48个百分点。

, authors=邵智超1, 邱杰凡1, 戴光麟1, 3, *, 李乔宏1, 陈翰墨1, 肖克江2, authorsList=邵智超, 邱杰凡, 戴光麟, 李乔宏, 陈翰墨, 肖克江, authorCompany=null, correspAuthors=戴光麟, authorNote=

2001年生。硕士研究生,CCF学生会员。主要研究方向为目标检测、人工智能

1984年生。博士,副教授。CCF会员。主要研究方向为嵌入式操作系统、物联网、人工智能

1979年生。博士,教授。CCF会员。主要研究方向为智能计算、医学图像处理、大数据分析

2004年生。学士。主要研究方向为人工智能、目标检测

2001年生。学士。主要研究方向为机器视觉、人工智能

1984年生。博士,教授。主要研究方向为多源信息融合、人工智能、智能感知互联、物联网系统

, correspAuthorsNote=
戴光麟()
, copyrightStatement=版权所有 © 2026 计算机研究与发展编辑部, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=s/TSgPz7n4Ii22cM3Mhcqg==, magXml=rjt0aHILoTpmBW60HvE2hg==, pdfUrl=null, pdf=6VYZ7MEGtTk6zZ62Y/Flig==, pdfFileSize=6129218, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=dJ5Vomry1t0zmD+JTKOBfA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=6JoXEhtewpwmyzJjwfHwmA==, mapNumber=null, fund=null)}, authors=[Author(id=1329420294973928141, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=211123030114@zjut.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420295066202831, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420294973928141, language=EN, stringName=Zhichao Shao, firstName=Zhichao, middleName=null, lastName=Shao, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, bio={"content":"

born in 2001. Master candidate. Student member of CCF. His main research interests include object detection and artificial intelligence

"}, bioImg=null, bioContent=

born in 2001. Master candidate. Student member of CCF. His main research interests include object detection and artificial intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420295133311696, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420294973928141, language=CN, stringName=邵智超, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1浙江工业大学计算机科学与技术学院 杭州 310023, bio={"img":"Bo9EM/4xkJRJ0XwIEuk72g==","content":"

2001年生。硕士研究生,CCF学生会员。主要研究方向为目标检测、人工智能

"}, bioImg=Bo9EM/4xkJRJ0XwIEuk72g==, bioContent=

2001年生。硕士研究生,CCF学生会员。主要研究方向为目标检测、人工智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)])]), Author(id=1329420295213003474, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420295288500948, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295213003474, language=EN, stringName=Jiefan Qiu, firstName=Jiefan, middleName=null, lastName=Qiu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, bio={"content":"

born in 1984. PhD, associate professor. Member of CCF. His main research interests include embedded operating systems, Internet of things, and artificial intelligence

"}, bioImg=null, bioContent=

born in 1984. PhD, associate professor. Member of CCF. His main research interests include embedded operating systems, Internet of things, and artificial intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420295351415509, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295213003474, language=CN, stringName=邱杰凡, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1浙江工业大学计算机科学与技术学院 杭州 310023, bio={"img":"AZggZSElLYgLeyzEOTu8+w==","content":"

1984年生。博士,副教授。CCF会员。主要研究方向为嵌入式操作系统、物联网、人工智能

"}, bioImg=AZggZSElLYgLeyzEOTu8+w==, bioContent=

1984年生。博士,副教授。CCF会员。主要研究方向为嵌入式操作系统、物联网、人工智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)])]), Author(id=1329420295418524375, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=dgl@zjut.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1329420295523381978, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295418524375, language=EN, stringName=Guanglin Dai, firstName=Guanglin, middleName=null, lastName=Dai, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, *, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023
3Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053, bio={"content":"

born in 1979. PhD, professor. Member of CCF. His main research interests include intelligent computing, medical image processing, and big data analysis

"}, bioImg=null, bioContent=

born in 1979. PhD, professor. Member of CCF. His main research interests include intelligent computing, medical image processing, and big data analysis

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420295590490843, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295418524375, language=CN, stringName=戴光麟, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, *, address=1浙江工业大学计算机科学与技术学院 杭州 310023
3浙江机电职业技术大学 杭州 310053, bio={"img":"CtQgNlyLFFzexIkIKMzQlw==","content":"

1979年生。博士,教授。CCF会员。主要研究方向为智能计算、医学图像处理、大数据分析

"}, bioImg=CtQgNlyLFFzexIkIKMzQlw==, bioContent=

1979年生。博士,教授。CCF会员。主要研究方向为智能计算、医学图像处理、大数据分析

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)]), AuthorCompany(id=1329420294894236361, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=3, ext=[AuthorCompanyExt(id=1329420294902624970, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294894236361, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053), AuthorCompanyExt(id=1329420294906819275, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294894236361, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3浙江机电职业技术大学 杭州 310053)])]), Author(id=1329420295678571229, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, 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=1329420295754068703, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295678571229, language=EN, stringName=Qiaohong Li, firstName=Qiaohong, middleName=null, lastName=Li, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, bio={"content":"

born in 2004. Bachelor. His main research interests include artificial intelligence and object detection

"}, bioImg=null, bioContent=

born in 2004. Bachelor. His main research interests include artificial intelligence and object detection

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420295812788960, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295678571229, language=CN, stringName=李乔宏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1浙江工业大学计算机科学与技术学院 杭州 310023, bio={"img":"3dh02iOZCClUFc87jQyYZA==","content":"

2004年生。学士。主要研究方向为人工智能、目标检测

"}, bioImg=3dh02iOZCClUFc87jQyYZA==, bioContent=

2004年生。学士。主要研究方向为人工智能、目标检测

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)])]), Author(id=1329420295879897826, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420295997338340, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295879897826, language=EN, stringName=Hanmo Chen, firstName=Hanmo, middleName=null, lastName=Chen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, bio={"content":"

born in 2001. Bachelor. His main research interests include machine vision and artificial intelligence

"}, bioImg=null, bioContent=

born in 2001. Bachelor. His main research interests include machine vision and artificial intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420296085418725, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295879897826, language=CN, stringName=陈翰墨, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1浙江工业大学计算机科学与技术学院 杭州 310023, bio={"img":"CIaN2u5Ao8tyA1dtq61Eyw==","content":"

2001年生。学士。主要研究方向为机器视觉、人工智能

"}, bioImg=CIaN2u5Ao8tyA1dtq61Eyw==, bioContent=

2001年生。学士。主要研究方向为机器视觉、人工智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)])]), Author(id=1329420296144138983, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=5, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420296215442153, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420296144138983, language=EN, stringName=Kejiang Xiao, firstName=Kejiang, middleName=null, lastName=Xiao, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Faculty of Artificial Intelligence in Education, Central China Normal University, Wuhan 430079, bio={"content":"

born in 1984. PhD, professor. His main research interests include multi-source information fusion, artificial intelligence, intelligent perception interconnection, and Internet of things systems

"}, bioImg=null, bioContent=

born in 1984. PhD, professor. His main research interests include multi-source information fusion, artificial intelligence, intelligent perception interconnection, and Internet of things systems

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420296290939626, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420296144138983, language=CN, stringName=肖克江, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2华中师范大学人工智能教育学部 武汉 430079, bio={"img":"IpVVR5ELuYkY8hdScN89GQ==","content":"

1984年生。博士,教授。主要研究方向为多源信息融合、人工智能、智能感知互联、物联网系统

"}, bioImg=IpVVR5ELuYkY8hdScN89GQ==, bioContent=

1984年生。博士,教授。主要研究方向为多源信息融合、人工智能、智能感知互联、物联网系统

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294806155974, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=2, ext=[AuthorCompanyExt(id=1329420294814544583, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294806155974, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Faculty of Artificial Intelligence in Education, Central China Normal University, Wuhan 430079), AuthorCompanyExt(id=1329420294822933192, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294806155974, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2华中师范大学人工智能教育学部 武汉 430079)])])], keywords=[Keyword(id=1329420296391602923, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, orderNo=1, keyword=object recognition), Keyword(id=1329420296458711788, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, orderNo=2, keyword=3D point cloud detection), Keyword(id=1329420298379703022, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, orderNo=3, keyword=machine vision), Keyword(id=1329420298442617584, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, orderNo=4, keyword=depth camera), Keyword(id=1329420298534892273, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, orderNo=5, keyword=plane detection), Keyword(id=1329420298597806834, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, orderNo=1, keyword=物品识别), Keyword(id=1329420298656527091, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, orderNo=2, keyword=3D点云探测), Keyword(id=1329420298732024564, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, orderNo=3, keyword=机器视觉), Keyword(id=1329420298790744821, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, orderNo=4, keyword=深度摄像头), Keyword(id=1329420298857853686, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, orderNo=5, keyword=平面检测)], refs=[Reference(id=1329420303412867893, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=null, journalName=null, refType=null, unstructuredReference=Park J, An G, Lee B N, et al. Real-time CNN-based object detection of prohibited items for X-ray security screening[J]. Radiation Physics and Chemistry, 2025, 232: 112681, articleTitle=null, refAbstract=null), Reference(id=1329420303484171062, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=Zu Mengmeng, Jia Jia, Wang Yuan, et al. Research on the application of target detection and recognition technology based on deep learning in industrial automation[C]//Proc of the 2nd Int Conf on Power, Electrical Engineering, Electronics and Control. Piscataway, NJ: IEEE, 2024: 615−620, articleTitle=null, refAbstract=null), Reference(id=1329420303563862839, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=Pei Jiabin, Wu Xiaoming, Liu Xiangzhi, et al. SGD-YOLOv5: A small object detection model for complex industrial environments[C]//Proc of the 41st Int Joint Conf on Neural Networks. Piscataway, NJ: IEEE, 2024: 1−10, articleTitle=null, refAbstract=null), Reference(id=1329420303639360312, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=3, authorNames=null, journalName=null, refType=null, unstructuredReference=Li Shiyang, Chen Jing, Peng Weimin, et al. A vehicle detection method based on disparity segmentation[J]. Multimedia Tools and Applications, 2023, 82(13): 19643−19655, articleTitle=null, refAbstract=null), Reference(id=1329420303719052089, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Yiqing. An analysis of the Viola-Jones face detection algorithm[J]. Image Processing on Line, 2014, 4: 128−148, articleTitle=null, refAbstract=null), Reference(id=1329420303790355258, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=Hassan M M, Hussein H I, Eesa A S, et al. Face recognition based on gabor feature extraction followed by fastica and LDA[J]. Computers, Materials and Continua, 2021, 68(2): 1637−1659, articleTitle=null, refAbstract=null), Reference(id=1329420303865852731, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=Ramu G, Babu S B G T. Image forgery detection for high resolution images using SIFT and RANSAC algorithm[C]//Proc of the 2nd Int Conf on Communication and Electronics Systems. Piscataway, NJ: IEEE, 2017: 850−854, articleTitle=null, refAbstract=null), Reference(id=1329420303937155900, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=Ravi N, Gabeur V, Hu Yuanting, et al. SAM2: Segment anything in images and videos[J]. arXiv preprint, arXiv: 2408.00714, 2024, articleTitle=null, refAbstract=null), Reference(id=1329420304008459069, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=Khanam R, Hussain M. YOLOv11: An overview of the key architectural enhancements[J]. arXiv preprint, arXiv: 2410.17725, 2024, articleTitle=null, refAbstract=null), Reference(id=1329420304071373630, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Honyi, Tseng Y, Chang K, et al. A denoising FPN with transformer R-CNN for tiny object detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 1−15, articleTitle=null, refAbstract=null), Reference(id=1329420304138482495, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhu Ran, Chen Silu, Li Xiaocong, et al. Recognition and localization method for high-speed robotic sorting in complex environments based on image interference assessment[J]. International Journal of Intelligent Robotics and Applications, 2025, 9(3): 1173−1186, articleTitle=null, refAbstract=null), Reference(id=1329420304201397056, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=Sidehabi S W, Azis M F, Asbar M. Development of a color-based image recognition system for robotic sorting and picking[J]. INTEK: Jurnal Penelitian, 2024, 11(2): 77−82, articleTitle=null, refAbstract=null), Reference(id=1329420304264311617, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=Titu M F S, Haque S M R, Islam R, et al. Experiments with cooperative robots that can detect object’s shape, color and size to perform tasks in industrial workplaces[J]. International Journal of Intelligent Robotics and Applications, 2024, 8(1): 179−192, articleTitle=null, refAbstract=null), Reference(id=1329420304339809090, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=Ali S, Pandey A K. ArthroNet: A monocular depth estimation technique with 3D segmented maps for knee arthroscopy[J]. Intelligent Medicine, 2023, 3(2): 129−138, articleTitle=null, refAbstract=null), Reference(id=1329420304436278083, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Sixing, Liu Mingmiao, Chai Yan, et al. Recognition and location of pepper picking based on improved YOLOv5s and depth camera[J]. Applied Engineering in Agriculture, 2023, 39(2): 179−185, articleTitle=null, refAbstract=null), Reference(id=1329420304499192644, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=Thanh P T H, Bui M Q V, Nguyen D D, et al. Transfer multi-source knowledge via scale-aware online domain adaptation in depth estimation for autonomous driving[J]. Image and Vision Computing, 2024, 141: 104871, articleTitle=null, refAbstract=null), Reference(id=1329420304562107205, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=Yang Tao, Gu Feifei. Overview of modulation techniques for spatially structured-light 3D imaging[J]. Optics & Laser Technology, 2024, 169: 110037, articleTitle=null, refAbstract=null), Reference(id=1329420304650187590, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=Zheng Yuanzhou, Liu Peng, Qian Long, et al. Recognition and depth estimation of ships based on binocular stereo vision[J]. Journal of Marine Science and Engineering, 2022, 10(8): 1153, articleTitle=null, refAbstract=null), Reference(id=1329420304725685063, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhang Hongxin, Lee S. Advancing the robotic vision revolution: Development and evaluation of a bionic binocular system for enhanced robotic vision[J]. Biomimetics, 2024, 9(6): 371, articleTitle=null, refAbstract=null), Reference(id=1329420304801182536, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=Bamji C, Godbaz J, Oh M, et al. A review of indirect time-of-flight technologies[J]. IEEE Transactions on Electron Devices, 2022, 69(6): 2779−2793, articleTitle=null, refAbstract=null), Reference(id=1329420304864097097, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=Paredes A L, Song Qiang, Conde M H. Performance evaluation of state-of-the-art high-resolution time-of-flight cameras[J]. IEEE Sensors Journal, 2023, 23(12): 13711−13727, articleTitle=null, refAbstract=null), Reference(id=1329420304927011658, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=Cao Xin, Wang Haoyu, Zhu Qiuquan, et al. PointStaClu: A deep point cloud clustering method based on stable cluster discrimination[J]. Remote Sensing, 2024, 16(13): 2423, articleTitle=null, refAbstract=null), Reference(id=1329420305002509131, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=22, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Jingyi, Liu Yu, Tan Hanlin, et al. A survey on weakly supervised 3D point cloud semantic segmentation[J]. IET Computer Vision, 2024, 18(3): 329−342, articleTitle=null, refAbstract=null), Reference(id=1329420305090589516, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=23, authorNames=null, journalName=null, refType=null, unstructuredReference=Noichl F, Collins F C, Braun A, et al. Enhancing point cloud semantic segmentation in the data-scarce domain of industrial plants through synthetic data[J]. Computer-Aided Civil and Infrastructure Engineering, 2024, 39(10): 1530−1549, articleTitle=null, refAbstract=null), Reference(id=1329420305153504077, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=Singh D P, Yadav M. Deep learning-based semantic segmentation of three-dimensional point cloud: A comprehensive review[J]. International Journal of Remote Sensing, 2024, 45(2): 532−586, articleTitle=null, refAbstract=null), Reference(id=1329420305224807246, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=26, rfOrder=25, authorNames=null, journalName=null, refType=null, unstructuredReference=韩冰, 张鑫云, 任爽. 基于三维点云的卷积运算综述[J]. 计算机研究与发展, 2023, 60(4): 873−902, articleTitle=null, refAbstract=null), Reference(id=1329420305287721807, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=26, rfOrder=26, authorNames=null, journalName=null, refType=null, unstructuredReference=Han Bing, Zhang Xinyun, Ren Shuang. Survey of convolution operations based on 3D point clouds[J]. Journal of Computer Research and Development, 2023, 60(4): 873−902 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1329420305350636368, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=27, rfOrder=27, authorNames=null, journalName=null, refType=null, unstructuredReference=吕卓, 姚治成, 贾玉祥, 等. 3D物体检测的异构方法[J]. 计算机研究与发展, 2021, 58(12): 2748−2759, articleTitle=null, refAbstract=null), Reference(id=1329420305417745233, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=27, rfOrder=28, authorNames=null, journalName=null, refType=null, unstructuredReference=Lü Zhuo, Yao Zhicheng, Jia Yuxiang, et al. A heterogeneous approach for 3D object detection[J]. Journal of Computer Research and Development, 2021, 58(12): 2748−2759 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1329420305480659794, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=28, rfOrder=29, authorNames=null, journalName=null, refType=null, unstructuredReference=廖瑞杰, 杨绍发, 孟文霞, 等. SegGraph: 室外场景三维点云闭环检测算法[J]. 计算机研究与发展, 2019, 56(2): 338−348, articleTitle=null, refAbstract=null), Reference(id=1329420305547768659, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=28, rfOrder=30, authorNames=null, journalName=null, refType=null, unstructuredReference=Liao Ruijie, Yang Shaofa, Meng Wenxia, et al. SegGraph: An algorithm for loop-closure detection in outdoor scenes using 3D point clouds[J]. Journal of Computer Research and Development, 2019, 56(2): 338−348 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1329420305606488916, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=29, rfOrder=31, authorNames=null, journalName=null, refType=null, unstructuredReference=Zeng Wenhao, Su Gongbing, Su Zixuan, et al. Research on point cloud registration and stitching fusion algorithm based on GCN-PRFNet[J]. IEEE Access, 2025, 13: 43384−43397, articleTitle=null, refAbstract=null), Reference(id=1329420305690374997, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=30, rfOrder=32, authorNames=null, journalName=null, refType=null, unstructuredReference=Eggen K, Fan Hongchao. WAPooling: An adaptive plug-and-play module for feature aggregation in point cloud classification networks[J]. International Journal of Applied Earth Observation and Geoinformation, 2025, 138: 104439, articleTitle=null, refAbstract=null), Reference(id=1329420305770066774, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=31, rfOrder=33, authorNames=null, journalName=null, refType=null, unstructuredReference=Zheng Yufei, Lu Jian, Chen Xiaogai, et al. MATNet: Semantic segmentation of 3D point clouds with multiscale adaptive transformer[J]. Computers and Electrical Engineering, 2024, 119: 109526, articleTitle=null, refAbstract=null), Reference(id=1329420305837175639, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=32, rfOrder=34, authorNames=null, journalName=null, refType=null, unstructuredReference=Li Chongshou, Liu Yuheng, Li Xinke, et al. Deep hierarchical learning for 3D semantic segmentation[J]. International Journal of Computer Vision, 2025, 133(7): 4420−4441, articleTitle=null, refAbstract=null), Reference(id=1329420305954616152, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=33, rfOrder=35, authorNames=null, journalName=null, refType=null, unstructuredReference=Wu Qiaoyun, Zhang Quanxiao, Tan Chunyu, et al. Point-to-spike residual learning for energy-efficient 3D point cloud classification[C]//Proc of the 38th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2024, 6092−6099, articleTitle=null, refAbstract=null), Reference(id=1329420306059473753, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=34, rfOrder=36, authorNames=null, journalName=null, refType=null, unstructuredReference=Holz D, Holzer S, Rusu R B, et al. Real-time plane segmentation using RGB-D cameras[C]//Proc of the 15th Robot Soccer World Cup. Berlin: Springer, 2011: 306−317, articleTitle=null, refAbstract=null), Reference(id=1329420306130776922, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=35, rfOrder=37, authorNames=null, journalName=null, refType=null, unstructuredReference=Poppinga J, Vaskevicius N, Birk A, et al. Fast plane detection and polygonalization in noisy 3D range images[C]//Proc of the 21st IEEE/RSJ Int Conf on Intelligent Robots and Systems. Piscataway, NJ: IEEE, 2008: 3378−3383, articleTitle=null, refAbstract=null), Reference(id=1329420306193691483, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=36, rfOrder=38, authorNames=null, journalName=null, refType=null, unstructuredReference=Jin Zhi, Tillo T, Zou Wenbin, et al. Robust plane detection using depth information from a consumer depth camera[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2017, 29(2): 447−460, articleTitle=null, refAbstract=null), Reference(id=1329420306298549084, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=37, rfOrder=39, authorNames=null, journalName=null, refType=null, unstructuredReference=El-Sayed E, Abdel-Kader R F, Nashaat H, et al. Plane detection in 3D point cloud using octree-balanced density down-sampling and iterative adaptive plane extraction[J]. IET Image Processing, 2018, 12(9): 1595−1605, articleTitle=null, refAbstract=null), Reference(id=1329420306369852253, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=38, rfOrder=40, authorNames=null, journalName=null, refType=null, unstructuredReference=Mouzenidis P, Louros A, Konstantinidis D, et al. Multi-modal variational faster R-CNN for improved visual object detection in manufacturing[C]//Proc of the 18th IEEE/CVF Int Conf on Computer Vision. Piscataway, NJ: IEEE, 2021: 2587−2594, articleTitle=null, refAbstract=null), Reference(id=1329420306445349726, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=39, rfOrder=41, authorNames=null, journalName=null, refType=null, unstructuredReference=Song Guozhu, Wang Jian, Ma Rongting, et al. Study on the fusion of improved YOLOv8 and depth camera for bunch tomato stem picking point recognition and localization[J]. Frontiers in Plant Science, 2024, 15: 1447855, articleTitle=null, refAbstract=null), Reference(id=1329420306533430111, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=40, rfOrder=42, authorNames=null, journalName=null, refType=null, unstructuredReference=ABB Robotics. PickMaster twin technical specification[EB/OL]. [2026-01-26]. https://www.abb.com/global/en/areas/robotics/products/software/pickmaster-twin, articleTitle=null, refAbstract=null), Reference(id=1329420306604733280, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=41, rfOrder=43, authorNames=null, journalName=null, refType=null, unstructuredReference=Lim S, Yoo Y, Lee J K, et al. Multi-object RANSAC: Efficient plane clustering method in a clutter[C]//Proc of the 41st IEEE Int Conf on Robotics and Automation. Piscataway, NJ: IEEE, 2024: 3079−3085, articleTitle=null, refAbstract=null), Reference(id=1329420306671842145, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=42, rfOrder=44, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Chen, Kim K, Gu Jinwei, et al. PlaneRCNN: 3D plane detection and reconstruction from a single image[C]//Proc of the 30th IEEE/CVF Conf on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2019: 4450−4459, articleTitle=null, refAbstract=null), Reference(id=1329420306734756706, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=43, rfOrder=45, authorNames=null, journalName=null, refType=null, unstructuredReference=Barath D, Matas J, Noskova J. MAGSAC: Marginalizing sample consensus[C]//Proc of the 30th IEEE/CVF Conf on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2019: 10197−10205, articleTitle=null, refAbstract=null), Reference(id=1329420306797671267, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=44, rfOrder=46, authorNames=null, journalName=null, refType=null, unstructuredReference=Liu Chen, Yang Jimei, Ceylan D, et al. PlaneNet: Piece-wise planar reconstruction from a single rgb image[C]//Proc of the 31st IEEE Conf on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2018: 2579−2588, articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)]), AuthorCompany(id=1329420294806155974, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=2, ext=[AuthorCompanyExt(id=1329420294814544583, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294806155974, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Faculty of Artificial Intelligence in Education, Central China Normal University, Wuhan 430079), AuthorCompanyExt(id=1329420294822933192, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294806155974, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2华中师范大学人工智能教育学部 武汉 430079)]), AuthorCompany(id=1329420294894236361, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=3, ext=[AuthorCompanyExt(id=1329420294902624970, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294894236361, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053), AuthorCompanyExt(id=1329420294906819275, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294894236361, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3浙江机电职业技术大学 杭州 310053)])], figs=[ArticleFig(id=1329420298992071415, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 1, caption=Sorting robot sorting scenario, figureFileSmall=fFuiZMr9bD31b2qtbXy/0Q==, figureFileBig=dJ5Vomry1t0zmD+JTKOBfA==, tableContent=null), ArticleFig(id=1329420299054985976, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图1, caption=分拣机器人分拣场景, figureFileSmall=fFuiZMr9bD31b2qtbXy/0Q==, figureFileBig=dJ5Vomry1t0zmD+JTKOBfA==, tableContent=null), ArticleFig(id=1329420299243729657, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 2, caption=Initial point cloud, figureFileSmall=c/uQxNBdvOsEf+0OYOdoKg==, figureFileBig=/x2VTf9ljDYde0iX6akEAA==, tableContent=null), ArticleFig(id=1329420299319227130, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图2, caption=初始点云, figureFileSmall=c/uQxNBdvOsEf+0OYOdoKg==, figureFileBig=/x2VTf9ljDYde0iX6akEAA==, tableContent=null), ArticleFig(id=1329420299390530299, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 3, caption=Incorrect plane fitting, figureFileSmall=I7G+NnJRJ0Txtq3PI4NBJQ==, figureFileBig=2/t6uLrqJTpjvpD0cA0enw==, tableContent=null), ArticleFig(id=1329420299474416380, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图3, caption=错误的平面拟合, figureFileSmall=I7G+NnJRJ0Txtq3PI4NBJQ==, figureFileBig=2/t6uLrqJTpjvpD0cA0enw==, tableContent=null), ArticleFig(id=1329420299554108157, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 4, caption=Overall flowchart of system, figureFileSmall=SPFmZGpORXDfgDMbXElGtg==, figureFileBig=U1RfQr9zRLbwQ8wmsoiAeA==, tableContent=null), ArticleFig(id=1329420299633799935, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图4, caption=系统整体流程图, figureFileSmall=SPFmZGpORXDfgDMbXElGtg==, figureFileBig=U1RfQr9zRLbwQ8wmsoiAeA==, tableContent=null), ArticleFig(id=1329420299759629056, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 5, caption=Initial scenario, figureFileSmall=gWu4V1oY4nFR1d6ZSJlXRw==, figureFileBig=P7ruHMkRzkpstb4uZsyp3Q==, tableContent=null), ArticleFig(id=1329420299822543617, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图5, caption=初始场景, figureFileSmall=gWu4V1oY4nFR1d6ZSJlXRw==, figureFileBig=P7ruHMkRzkpstb4uZsyp3Q==, tableContent=null), ArticleFig(id=1329420299889652482, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 6, caption=Removal effect for point cloud of plane that disturbance objects located on, figureFileSmall=UrwVYzLOuJ+mEb0Ffk48oA==, figureFileBig=aQCGSis+Z3SCbwu3WZLVoQ==, tableContent=null), ArticleFig(id=1329420299952567043, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图6, caption=干扰物所处平面点云去除效果, figureFileSmall=UrwVYzLOuJ+mEb0Ffk48oA==, figureFileBig=aQCGSis+Z3SCbwu3WZLVoQ==, tableContent=null), ArticleFig(id=1329420300023870212, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 7, caption=Plane and overall point cloud, figureFileSmall=EbWZ9TrH/lgGzpMtRgWGnA==, figureFileBig=vcS3N+XR/tixfZE0Vz23LQ==, tableContent=null), ArticleFig(id=1329420300082590469, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图7, caption=平面与整体点云, figureFileSmall=EbWZ9TrH/lgGzpMtRgWGnA==, figureFileBig=vcS3N+XR/tixfZE0Vz23LQ==, tableContent=null), ArticleFig(id=1329420300149699334, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 8, caption=Visualization of convex hull calculation and coordinate mapping results, figureFileSmall=TAbmi8wYgtj/ID6SaUmbiQ==, figureFileBig=l9aDcvXXjfffAnjURqJJ2g==, tableContent=null), ArticleFig(id=1329420300225196807, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图8, caption=凸包计算与坐标映射结果可视化, figureFileSmall=TAbmi8wYgtj/ID6SaUmbiQ==, figureFileBig=l9aDcvXXjfffAnjURqJJ2g==, tableContent=null), ArticleFig(id=1329420300288111368, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 9, caption=Visualization diagram of point cloud detection partial results, figureFileSmall=o4HXNf5yPtTC2s2/0aQMEQ==, figureFileBig=Vy8s6OBRD7/tm4PpHIZgwg==, tableContent=null), ArticleFig(id=1329420300342637321, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图9, caption=点云探测部分结果可视化图, figureFileSmall=o4HXNf5yPtTC2s2/0aQMEQ==, figureFileBig=Vy8s6OBRD7/tm4PpHIZgwg==, tableContent=null), ArticleFig(id=1329420300401357578, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 10, caption=Experimental scenario layout and experimental devices, figureFileSmall=ZUShdV1xuqxpNYys+RtFOA==, figureFileBig=YoKk60CVSStSIp3i9d4UXQ==, tableContent=null), ArticleFig(id=1329420300460077836, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图10, caption=实验场景布置及实验设备, figureFileSmall=ZUShdV1xuqxpNYys+RtFOA==, figureFileBig=YoKk60CVSStSIp3i9d4UXQ==, tableContent=null), ArticleFig(id=1329420300535575309, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 11, caption=Specific experimental scenarios, figureFileSmall=1pIxRmO+07rhYGleGBkQxw==, figureFileBig=I5YtDZL9mxTq8/ETGbPHEg==, tableContent=null), ArticleFig(id=1329420300627849998, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图11, caption=具体实验场景, figureFileSmall=1pIxRmO+07rhYGleGBkQxw==, figureFileBig=I5YtDZL9mxTq8/ETGbPHEg==, tableContent=null), ArticleFig(id=1329420300711736079, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 12, caption=Specific multi-desktop experimental scenarios, figureFileSmall=Xsv/WYXA1wAzH6oeHWy06w==, figureFileBig=ZWLx9XLcxwUUxpu/HzMNpg==, tableContent=null), ArticleFig(id=1329420300774650640, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图12, caption=具体多桌面实验场景, figureFileSmall=Xsv/WYXA1wAzH6oeHWy06w==, figureFileBig=ZWLx9XLcxwUUxpu/HzMNpg==, tableContent=null), ArticleFig(id=1329420300833370897, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 13, caption=Performance comparison of 3Deus and other algorithms, figureFileSmall=2KrPnwcEAxvDyAGHmyRgAg==, figureFileBig=HrkUXQKipGAme92Op3u4gQ==, tableContent=null), ArticleFig(id=1329420300887896850, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图13, caption=3Deus与其他算法性能对比, figureFileSmall=2KrPnwcEAxvDyAGHmyRgAg==, figureFileBig=HrkUXQKipGAme92Op3u4gQ==, tableContent=null), ArticleFig(id=1329420300946617107, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 14, caption=Recognition results for object on the table, figureFileSmall=3lQHlqcfDjg6YpkIUMxBYA==, figureFileBig=47OOPMkaJ5KgpGqVdZrSUQ==, tableContent=null), ArticleFig(id=1329420301013725972, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图14, caption=桌上物品识别结果, figureFileSmall=3lQHlqcfDjg6YpkIUMxBYA==, figureFileBig=47OOPMkaJ5KgpGqVdZrSUQ==, tableContent=null), ArticleFig(id=1329420301106000661, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 15, caption=Recognition results for objects under the table, figureFileSmall=kWgFfaZVaFYwXn2dr/qaHQ==, figureFileBig=c+VUWkbo7Wfs6fqTRqt3Tg==, tableContent=null), ArticleFig(id=1329420301168915222, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图15, caption=桌下物品识别结果, figureFileSmall=kWgFfaZVaFYwXn2dr/qaHQ==, figureFileBig=c+VUWkbo7Wfs6fqTRqt3Tg==, tableContent=null), ArticleFig(id=1329420301227635479, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 16, caption=Recognition results for objects beside the table, figureFileSmall=phkuTzjpDEztUPdC+6NKeA==, figureFileBig=PHLGmNreMytjHstFZI1YFQ==, tableContent=null), ArticleFig(id=1329420301298938648, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图16, caption=桌旁物品识别结果, figureFileSmall=phkuTzjpDEztUPdC+6NKeA==, figureFileBig=PHLGmNreMytjHstFZI1YFQ==, tableContent=null), ArticleFig(id=1329420301374436121, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 17, caption=Overall recognition results of each scenario, figureFileSmall=DfeEEanybmjAJ5OycKuOEg==, figureFileBig=BgavwFFgHDUuMvcYCYZnCQ==, tableContent=null), ArticleFig(id=1329420301424767770, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图17, caption=各场景整体识别结果, figureFileSmall=DfeEEanybmjAJ5OycKuOEg==, figureFileBig=BgavwFFgHDUuMvcYCYZnCQ==, tableContent=null), ArticleFig(id=1329420301483488027, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 18, caption=Recognition results under single-plane scenario, figureFileSmall=vYnxlTgAY0GBh1aBaCdSlQ==, figureFileBig=vaf9CQQYFgnTcd7dnaoM5w==, tableContent=null), ArticleFig(id=1329420301596734236, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图18, caption=单平面场景下的识别结果, figureFileSmall=vYnxlTgAY0GBh1aBaCdSlQ==, figureFileBig=vaf9CQQYFgnTcd7dnaoM5w==, tableContent=null), ArticleFig(id=1329420301668037405, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 19, caption=Recognition results under multi-plane complex scenarios, figureFileSmall=bS4HKY5ATnT8xA8SSNCdhQ==, figureFileBig=dwPSz1QwFHRSAEhYCd0vbA==, tableContent=null), ArticleFig(id=1329420301739340574, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图19, caption=多平面复杂场景下的识别结果, figureFileSmall=bS4HKY5ATnT8xA8SSNCdhQ==, figureFileBig=dwPSz1QwFHRSAEhYCd0vbA==, tableContent=null), ArticleFig(id=1329420301810643743, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 20, caption=Performance of the algorithm under the multi-horizontal desktop experimental scenario, figureFileSmall=KbIZOTIZdF15N5Lgpg5tyw==, figureFileBig=HQKZbeqxwVPpMhcxHYlHxA==, tableContent=null), ArticleFig(id=1329420301873558304, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图20, caption=多水平桌面实验场景下的算法性能, figureFileSmall=KbIZOTIZdF15N5Lgpg5tyw==, figureFileBig=HQKZbeqxwVPpMhcxHYlHxA==, tableContent=null), ArticleFig(id=1329420301932278561, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 21, caption=Performance of the algorithm under the horizontal plus inclined desktop experimental scenario, figureFileSmall=n7VvNS/XJ4BxJqBXzp+b7A==, figureFileBig=Pj/pgtF1ogg+L3TYncCN1A==, tableContent=null), ArticleFig(id=1329420301990998818, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图21, caption=水平加倾斜桌面实验场景下的算法性能, figureFileSmall=n7VvNS/XJ4BxJqBXzp+b7A==, figureFileBig=Pj/pgtF1ogg+L3TYncCN1A==, tableContent=null), ArticleFig(id=1329420302049719075, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Figure 22, caption=Performance of the algorithm under the uneven desktop experimental scenario, figureFileSmall=rrI2gitJhXRR5ffxhTJawA==, figureFileBig=O/Tslu0Eqh+Na+uiY+3aZg==, tableContent=null), ArticleFig(id=1329420302108439332, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=图22, caption=不平整桌面实验场景下的算法性能, figureFileSmall=rrI2gitJhXRR5ffxhTJawA==, figureFileBig=O/Tslu0Eqh+Na+uiY+3aZg==, tableContent=null), ArticleFig(id=1329420302179742501, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Table 1, caption=

Parameters of the Depth Camera

, figureFileSmall=null, figureFileBig=null, tableContent=
参数取值
深度信息精度/mm3
跟踪的对象的最大运动范围/m4
深度滤波级别6
可靠性阈值3
), ArticleFig(id=1329420302246851366, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=表1, caption=

深度摄像头参数

, figureFileSmall=null, figureFileBig=null, tableContent=
参数取值
深度信息精度/mm3
跟踪的对象的最大运动范围/m4
深度滤波级别6
可靠性阈值3
), ArticleFig(id=1329420302318154535, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Table 2, caption=

Confusion Matrix of 3Deus

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实被预测数量
桌上桌下桌旁
干扰场景1桌上49901
桌下040
桌旁000
干扰场景2桌上511120
桌下02130
桌旁1200
干扰场景3桌上519915
桌下01490
桌旁3091
干扰场景4桌上4701223
桌下01860
桌旁20108
), ArticleFig(id=1329420302402040616, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=表2, caption=

3Deus混淆矩阵

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实被预测数量
桌上桌下桌旁
干扰场景1桌上49901
桌下040
桌旁000
干扰场景2桌上511120
桌下02130
桌旁1200
干扰场景3桌上519915
桌下01490
桌旁3091
干扰场景4桌上4701223
桌下01860
桌旁20108
), ArticleFig(id=1329420302473343785, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Table 3, caption=

Confusion Matrix Only Removing Ground Point Cloud Algorithm

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实类别
桌上桌下桌旁
干扰场景1桌上49801
桌下030
桌旁110
干扰场景2桌上515110
桌下02080
桌旁860
干扰场景3桌上52111106
桌下01460
桌旁110
干扰场景4桌上47214132
桌下01830
桌旁011
), ArticleFig(id=1329420302544646954, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=表3, caption=

仅去除地面点云算法的混淆矩阵

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实类别
桌上桌下桌旁
干扰场景1桌上49801
桌下030
桌旁110
干扰场景2桌上515110
桌下02080
桌旁860
干扰场景3桌上52111106
桌下01460
桌旁110
干扰场景4桌上47214132
桌下01830
桌旁011
), ArticleFig(id=1329420302611755819, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Table 4, caption=

Recognition Confusion Matrix Without Plane Expansion Mechanism

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实类别
桌上桌下桌旁
干扰场景1桌上40001
桌下040
桌旁9900
干扰场景2桌上37670
桌下02170
桌旁14710
干扰场景3桌上368107
桌下01472
桌旁154197
干扰场景4桌上31467
桌下01865
桌旁1586119
), ArticleFig(id=1329420302691447596, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=表4, caption=

无平面扩张机制的识别混淆矩阵

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实类别
桌上桌下桌旁
干扰场景1桌上40001
桌下040
桌旁9900
干扰场景2桌上37670
桌下02170
桌旁14710
干扰场景3桌上368107
桌下01472
桌旁154197
干扰场景4桌上31467
桌下01865
桌旁1586119
), ArticleFig(id=1329420302762750765, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Table 5, caption=

Confusion Matrix of 3Deus Under Multi-horizontal Desktop Experimental Scenarios

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实被预测数量
桌上桌下桌旁
干扰场景1桌上53500
桌下000
桌旁1000
干扰场景2桌上562160
桌下0490
桌旁730
干扰场景3桌上561422
桌下01210
桌旁140174
干扰场景4桌上541523
桌下01698
桌旁160189
), ArticleFig(id=1329420302834053934, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=表5, caption=

多水平桌面实验场景下3Deus的混淆矩阵

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实被预测数量
桌上桌下桌旁
干扰场景1桌上53500
桌下000
桌旁1000
干扰场景2桌上562160
桌下0490
桌旁730
干扰场景3桌上561422
桌下01210
桌旁140174
干扰场景4桌上541523
桌下01698
桌旁160189
), ArticleFig(id=1329420302905357103, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Table 6, caption=

Confusion Matrix of 3Deus Under Horizontal Plus Inclined Desktop Experimental Scenarios

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实被预测数量
桌上桌下桌旁
干扰场景1桌上50700
桌下800
桌旁600
干扰场景2桌上51800
桌下41540
桌旁1900
干扰场景3桌上51402
桌下61371
桌旁270106
干扰场景4桌上51305
桌下81030
桌旁271159
), ArticleFig(id=1329420302976660272, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=表6, caption=

水平加倾斜桌面实验场景下3Deus的混淆矩阵

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实被预测数量
桌上桌下桌旁
干扰场景1桌上50700
桌下800
桌旁600
干扰场景2桌上51800
桌下41540
桌旁1900
干扰场景3桌上51402
桌下61371
桌旁270106
干扰场景4桌上51305
桌下81030
桌旁271159
), ArticleFig(id=1329420303060546353, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Table 7, caption=

Confusion Matrix of 3Deus Under Uneven Desktop Experimental Scenarios

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实类别
桌上桌下桌旁
干扰场景1桌上55400
桌下100
桌旁1100
干扰场景2桌上54700
桌下01550
桌旁1600
干扰场景3桌上52727
桌下0630
桌旁102134
干扰场景4桌上57793
桌下01082
桌旁230129
), ArticleFig(id=1329420303169598258, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=表7, caption=

不平整桌面实验场景下3Deus的混淆矩阵

, figureFileSmall=null, figureFileBig=null, tableContent=
场景预测种类真实类别
桌上桌下桌旁
干扰场景1桌上55400
桌下100
桌旁1100
干扰场景2桌上54700
桌下01550
桌旁1600
干扰场景3桌上52727
桌下0630
桌旁102134
干扰场景4桌上57793
桌下01082
桌旁230129
), ArticleFig(id=1329420303240901427, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=EN, label=Table 8, caption=

Running Time on Different Devices s

, figureFileSmall=null, figureFileBig=null, tableContent=
设备名称平均运行时间
Jetson Nano12.56
Jetson NX4.24
PC机2.35
), ArticleFig(id=1329420303308010292, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, language=CN, label=表8, caption=

不同设备上的运行时间

, figureFileSmall=null, figureFileBig=null, tableContent=
设备名称平均运行时间
Jetson Nano12.56
Jetson NX4.24
PC机2.35
)], attaches=null, journal=Journal(id=1324759606947315908, delFlag=0, nameCn=计算机研究与发展, nameEn=Journal of Computer Research and Development, nameHistory1=null, nameHistory2=null, issn=1000-1239, eissn=null, cn=11-1777/TP, coden=null, periodic=0, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=k2l8zgqwstuxxevFX2qxVA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Computer Research and Development, journalRemark=null, publicationField=null, createdTime=1793776903141, updatedTime=1793850698612, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Electronic Information, subjectName=null, subjectCodeEn=Electronic Information, subjectNameEn=null, picCn=k2l8zgqwstuxxevFX2qxVA==, picEn=r6k0Ry+RD+c/VXj5QarmDw==, jcr=null, cjcr=null, exts=[JournalExt(id=1325069127710286688, language=CN, name=计算机研究与发展, 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=1793850698640, updatedTime=1793850698640, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://crad.ict.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://crad.ict.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://crad.ict.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=https://crad.ict.ac.cn/journalx/editorInChiefLogOn.action, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1325069127777395553, language=EN, name=Journal of Computer Research and Development, 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=1793850698656, updatedTime=1793850698656, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://crad.ict.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://crad.ict.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://crad.ict.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=https://crad.ict.ac.cn/journalx/editorInChiefLogOn.action, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1324759914146529478, websiteList=[Website(id=1325070300139892771, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1324759914146529478, 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/jsjyjyfz/CN, language=CN, createTime=1793850978169, createBy=18614031015, updateTime=1793851040259, updateBy=18614031015, name=计算机研究与发展-中文, tplId=1146099689490845704, title=计算机研究与发展, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1325070893751349360, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=articleTextType, value=kx, createTime=1793851119697, updateTime=1793851119697, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893730377837, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=banner, value=null, createTime=1793851119692, updateTime=1793851119692, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893772320883, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=grayFlag, value=0, createTime=1793851119702, updateTime=1793851119702, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893721989228, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=logo, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/file/pic?fileId=U27tDsy717YF7XSPA+Vwdw==, createTime=1793851119690, updateTime=1793851119690, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893784903797, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=minRunFlag, value=0, createTime=1793851119705, updateTime=1793851119705, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893742960751, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/file/pic, createTime=1793851119695, updateTime=1793851119695, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893780709492, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=silenceFlag, value=0, createTime=1793851119704, updateTime=1793851119704, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893738766446, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1793851119694, updateTime=1793851119694, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893755543665, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=themeColor, value=null, createTime=1793851119698, updateTime=1793851119698, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893763932274, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=themeStyle, value=null, createTime=1793851119700, updateTime=1793851119700, creator=18614031015, updator=18614031015)]), Website(id=1325070300580294698, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1324759914146529478, 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/jsjyjyfz/EN, language=EN, createTime=1793850978274, createBy=18614031015, updateTime=1793851036638, updateBy=18614031015, name=计算机研究与发展-英文, tplId=1146101810881728533, title=Journal of Computer Research and Development, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1325070925288321148, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=articleTextType, value=kx, createTime=1793851127216, updateTime=1793851127216, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925263155321, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=banner, value=null, createTime=1793851127210, updateTime=1793851127210, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925309292671, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=grayFlag, value=0, createTime=1793851127221, updateTime=1793851127221, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925254766712, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=logo, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/file/pic?fileId=U27tDsy717YF7XSPA+Vwdw==, createTime=1793851127208, updateTime=1793851127208, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925326069889, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=minRunFlag, value=0, createTime=1793851127225, updateTime=1793851127225, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925279932539, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/file/pic, createTime=1793851127214, updateTime=1793851127214, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925317681280, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=silenceFlag, value=0, createTime=1793851127223, updateTime=1793851127223, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925271543930, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1793851127212, updateTime=1793851127212, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925292515453, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=themeColor, value=null, createTime=1793851127217, updateTime=1793851127217, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925300904062, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=themeStyle, value=null, createTime=1793851127219, updateTime=1793851127219, creator=18614031015, updator=18614031015)])], journalTitle=计算机研究与发展, weixinUrl=null, journalUrl=https://crad.ict.ac.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Computer Research and Development, journalPhotoCn=k2l8zgqwstuxxevFX2qxVA==, journalPhotoEn=r6k0Ry+RD+c/VXj5QarmDw==, 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, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/10.7544/issn1000-1239.202550696, detailUrlEn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/10.7544/issn1000-1239.202550696, pdfUrlCn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/PDF/10.7544/issn1000-1239.202550696, pdfUrlEn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/PDF/10.7544/issn1000-1239.202550696, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1785513600000, fullTextJson=null, articleText=null, reference=null)
收藏切换
3Deus:基于点云探测的复杂场景物品识别方法
邵智超 1 , 邱杰凡 1 , 戴光麟 1, 3, * , 李乔宏 1 , 陈翰墨 1 , 肖克江 2
计算机研究与发展 | 人工智能 2026,63(8): 2076-2094
收起
收藏切换
计算机研究与发展 |人工智能 2026 , 63 (8) : 2076 -2094
3Deus:基于点云探测的复杂场景物品识别方法
全屏
[Author(id=1329420294973928141, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=211123030114@zjut.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420295066202831, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420294973928141, language=EN, stringName=Zhichao Shao, firstName=Zhichao, middleName=null, lastName=Shao, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, bio={"content":"

born in 2001. Master candidate. Student member of CCF. His main research interests include object detection and artificial intelligence

"}, bioImg=null, bioContent=

born in 2001. Master candidate. Student member of CCF. His main research interests include object detection and artificial intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420295133311696, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420294973928141, language=CN, stringName=邵智超, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1浙江工业大学计算机科学与技术学院 杭州 310023, bio={"img":"Bo9EM/4xkJRJ0XwIEuk72g==","content":"

2001年生。硕士研究生,CCF学生会员。主要研究方向为目标检测、人工智能

"}, bioImg=Bo9EM/4xkJRJ0XwIEuk72g==, bioContent=

2001年生。硕士研究生,CCF学生会员。主要研究方向为目标检测、人工智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)])]), Author(id=1329420295213003474, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420295288500948, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295213003474, language=EN, stringName=Jiefan Qiu, firstName=Jiefan, middleName=null, lastName=Qiu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, bio={"content":"

born in 1984. PhD, associate professor. Member of CCF. His main research interests include embedded operating systems, Internet of things, and artificial intelligence

"}, bioImg=null, bioContent=

born in 1984. PhD, associate professor. Member of CCF. His main research interests include embedded operating systems, Internet of things, and artificial intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420295351415509, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295213003474, language=CN, stringName=邱杰凡, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1浙江工业大学计算机科学与技术学院 杭州 310023, bio={"img":"AZggZSElLYgLeyzEOTu8+w==","content":"

1984年生。博士,副教授。CCF会员。主要研究方向为嵌入式操作系统、物联网、人工智能

"}, bioImg=AZggZSElLYgLeyzEOTu8+w==, bioContent=

1984年生。博士,副教授。CCF会员。主要研究方向为嵌入式操作系统、物联网、人工智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)])]), Author(id=1329420295418524375, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=dgl@zjut.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1329420295523381978, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295418524375, language=EN, stringName=Guanglin Dai, firstName=Guanglin, middleName=null, lastName=Dai, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, *, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023
3Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053, bio={"content":"

born in 1979. PhD, professor. Member of CCF. His main research interests include intelligent computing, medical image processing, and big data analysis

"}, bioImg=null, bioContent=

born in 1979. PhD, professor. Member of CCF. His main research interests include intelligent computing, medical image processing, and big data analysis

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420295590490843, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295418524375, language=CN, stringName=戴光麟, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 3, *, address=1浙江工业大学计算机科学与技术学院 杭州 310023
3浙江机电职业技术大学 杭州 310053, bio={"img":"CtQgNlyLFFzexIkIKMzQlw==","content":"

1979年生。博士,教授。CCF会员。主要研究方向为智能计算、医学图像处理、大数据分析

"}, bioImg=CtQgNlyLFFzexIkIKMzQlw==, bioContent=

1979年生。博士,教授。CCF会员。主要研究方向为智能计算、医学图像处理、大数据分析

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)]), AuthorCompany(id=1329420294894236361, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=3, ext=[AuthorCompanyExt(id=1329420294902624970, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294894236361, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053), AuthorCompanyExt(id=1329420294906819275, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294894236361, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3浙江机电职业技术大学 杭州 310053)])]), Author(id=1329420295678571229, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, 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=1329420295754068703, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295678571229, language=EN, stringName=Qiaohong Li, firstName=Qiaohong, middleName=null, lastName=Li, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, bio={"content":"

born in 2004. Bachelor. His main research interests include artificial intelligence and object detection

"}, bioImg=null, bioContent=

born in 2004. Bachelor. His main research interests include artificial intelligence and object detection

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420295812788960, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295678571229, language=CN, stringName=李乔宏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1浙江工业大学计算机科学与技术学院 杭州 310023, bio={"img":"3dh02iOZCClUFc87jQyYZA==","content":"

2004年生。学士。主要研究方向为人工智能、目标检测

"}, bioImg=3dh02iOZCClUFc87jQyYZA==, bioContent=

2004年生。学士。主要研究方向为人工智能、目标检测

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)])]), Author(id=1329420295879897826, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420295997338340, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295879897826, language=EN, stringName=Hanmo Chen, firstName=Hanmo, middleName=null, lastName=Chen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, bio={"content":"

born in 2001. Bachelor. His main research interests include machine vision and artificial intelligence

"}, bioImg=null, bioContent=

born in 2001. Bachelor. His main research interests include machine vision and artificial intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420296085418725, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420295879897826, language=CN, stringName=陈翰墨, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1浙江工业大学计算机科学与技术学院 杭州 310023, bio={"img":"CIaN2u5Ao8tyA1dtq61Eyw==","content":"

2001年生。学士。主要研究方向为机器视觉、人工智能

"}, bioImg=CIaN2u5Ao8tyA1dtq61Eyw==, bioContent=

2001年生。学士。主要研究方向为机器视觉、人工智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294718075587, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=1, ext=[AuthorCompanyExt(id=1329420294730658500, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023), AuthorCompanyExt(id=1329420294739047109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294718075587, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1浙江工业大学计算机科学与技术学院 杭州 310023)])]), Author(id=1329420296144138983, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, orderNo=5, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420296215442153, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420296144138983, language=EN, stringName=Kejiang Xiao, firstName=Kejiang, middleName=null, lastName=Xiao, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Faculty of Artificial Intelligence in Education, Central China Normal University, Wuhan 430079, bio={"content":"

born in 1984. PhD, professor. His main research interests include multi-source information fusion, artificial intelligence, intelligent perception interconnection, and Internet of things systems

"}, bioImg=null, bioContent=

born in 1984. PhD, professor. His main research interests include multi-source information fusion, artificial intelligence, intelligent perception interconnection, and Internet of things systems

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420296290939626, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, authorId=1329420296144138983, language=CN, stringName=肖克江, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2华中师范大学人工智能教育学部 武汉 430079, bio={"img":"IpVVR5ELuYkY8hdScN89GQ==","content":"

1984年生。博士,教授。主要研究方向为多源信息融合、人工智能、智能感知互联、物联网系统

"}, bioImg=IpVVR5ELuYkY8hdScN89GQ==, bioContent=

1984年生。博士,教授。主要研究方向为多源信息融合、人工智能、智能感知互联、物联网系统

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420294806155974, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, xref=2, ext=[AuthorCompanyExt(id=1329420294814544583, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294806155974, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Faculty of Artificial Intelligence in Education, Central China Normal University, Wuhan 430079), AuthorCompanyExt(id=1329420294822933192, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420287659061894, companyId=1329420294806155974, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2华中师范大学人工智能教育学部 武汉 430079)])])]
邵智超1 , 邱杰凡1, 戴光麟1, 3, * , 李乔宏1, 陈翰墨1, 肖克江2
作者信息
  • 1浙江工业大学计算机科学与技术学院 杭州 310023
  • 2华中师范大学人工智能教育学部 武汉 430079
  • 3浙江机电职业技术大学 杭州 310053
通讯作者:
戴光麟()
作者简介:

2001年生。硕士研究生,CCF学生会员。主要研究方向为目标检测、人工智能

1984年生。博士,副教授。CCF会员。主要研究方向为嵌入式操作系统、物联网、人工智能

1979年生。博士,教授。CCF会员。主要研究方向为智能计算、医学图像处理、大数据分析

2004年生。学士。主要研究方向为人工智能、目标检测

2001年生。学士。主要研究方向为机器视觉、人工智能

1984年生。博士,教授。主要研究方向为多源信息融合、人工智能、智能感知互联、物联网系统

3Deus: Object Recognition Method Based on Point Cloud Detection in Complex Scenario
Zhichao Shao1 , Jiefan Qiu1, Guanglin Dai1, 3, * , Qiaohong Li1, Hanmo Chen1, Kejiang Xiao2
Affiliations
  • 1College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023
  • 2Faculty of Artificial Intelligence in Education, Central China Normal University, Wuhan 430079
  • 3Zhejiang Polytechnic University of Mechanical and Electrical Engineering, Hangzhou 310053
  • born in 2001. Master candidate. Student member of CCF. His main research interests include object detection and artificial intelligence

    born in 1984. PhD, associate professor. Member of CCF. His main research interests include embedded operating systems, Internet of things, and artificial intelligence

    born in 1979. PhD, professor. Member of CCF. His main research interests include intelligent computing, medical image processing, and big data analysis

    born in 2004. Bachelor. His main research interests include artificial intelligence and object detection

    born in 2001. Bachelor. His main research interests include machine vision and artificial intelligence

    born in 1984. PhD, professor. His main research interests include multi-source information fusion, artificial intelligence, intelligent perception interconnection, and Internet of things systems

出版时间: 2026-08-01 doi: 10.7544/issn1000-1239.202550696
文章导航
收藏切换

随着人工智能技术的发展,基于视觉的物品识别作为构建物联网感知层的重要组成部分,与分拣机器人相结合后可应用在各类货物分拣的具体场景中。然而,在混乱摆放物品的复杂场景中,由于目标物品与干扰物品可能存在空间距离差,仅依赖物品识别算法并不能很好地区分目标物品和干扰物品,势必对机器人的正常抓取动作产生负面影响。为了解决该问题,提出了一种轻量化的复杂环境下的物品识别方法——3Deus。该方法将深度学习和3D点云数据处理相结合,利用3D点云对干扰物品所处的平面做整体消除,并基于改进的随机采样一致性算法区分目标物品所处平面与堆叠干扰物;运用平面扩张机制,保证目标物品所处平面点云的完整性;最后,借助物品识别算法获取目标的识别框并构建其与具体点云簇的映射关系。将3Deus部署在边缘计算设备上,并通过构建不同的实验场景验证了3Deus的有效性和鲁棒性。实验结果表明,与传统的随机采样一致性算法相比,3Deus能够有效地对目标物品和干扰物品进行区分,其准确率达到了95%以上,同时3Deus利用平面扩张机制将目标物品的识别准确率提高了30.48个百分点。

物品识别  /  3D点云探测  /  机器视觉  /  深度摄像头  /  平面检测

With the development of artificial intelligence technology, vision-based object recognition, as an important component of the Internet of things perception layer, can be applied in various cargo sorting scenarios when combined with sorting robots. However, in complex environments where objects are disorganized, relying solely on object recognition algorithms may not be sufficient to distinguish target objects from interfering ones, as the difference between them may be the spatial distance. This can negatively impact the robot’s normal grasping actions. To address this issue, we propose a lightweight object recognition method for complex environments—3Deus. This method combines deep learning with 3D point cloud data processing. It uses a depth camera to synchronously capture both RGB images and 3D point cloud data in the recognition scene. The method eliminates the plane where interfering objects are located using the 3D point cloud, and distinguishes the plane containing the target object from the stacked interfering objects with an improved RANSAC algorithm. It also employs a plane expansion mechanism to ensure the integrity of the plane point cloud where the target object is located. Finally, the object recognition algorithm is used to obtain the recognition bounding box of the target object and construct its mapping relationship with the specific point cloud cluster. We deploy 3Deus on edge computing devices and validate its effectiveness and robustness by constructing different experimental scenarios. Experimental results show that compared with the traditional random sampling consensus algorithm, 3Deus can effectively distinguish the target objects from interference objects, achieving an accuracy of over 95%. Moreover, by utilizing the plane expansion mechanism, the recognition accuracy of the target object is improved by 30.48%.

object recognition  /  3D point cloud detection  /  machine vision  /  depth camera  /  plane detection
邵智超, 邱杰凡, 戴光麟, 李乔宏, 陈翰墨, 肖克江. 3Deus:基于点云探测的复杂场景物品识别方法. 计算机研究与发展, 2026 , 63 (8) : 2076 -2094 . DOI: 10.7544/issn1000-1239.202550696
Zhichao Shao, Jiefan Qiu, Guanglin Dai, Qiaohong Li, Hanmo Chen, Kejiang Xiao. 3Deus: Object Recognition Method Based on Point Cloud Detection in Complex Scenario[J]. Journal of Computer Research and Development, 2026 , 63 (8) : 2076 -2094 . DOI: 10.7544/issn1000-1239.202550696
物品识别作为物联网感知层的一项重要任务,是物联网实现环境信息采集与智能处理的核心技术之一。近年来,随着基于视觉的物品识别算法的快速发展,其被广泛地运用在如违禁物品识别[1]、工业生产[2-3]、交通运输[4]等多个领域。目前,物品识别算法主要分为基于传统视觉模型与基于深度学习模型2类。基于传统视觉模型的物品识别算法主要包括Viola-Jones算法[5]、基于Gabor和LDA结合方法[6]、基于SIFT特征和RANSAC匹配算法[7]等。传统模型往往基于手工特征提取,使用场景有限。相反,深度学习模型利用自身的特征提取器,使其具有较高的鲁棒性。常用的深度学习模型包括SAM2[8],YOLOv11[9],DN-FPN[10]等。
然而,随着工业物联网应用场景中(需要结合空间信息的场景)对物品识别复杂度的不断增加,单一的物品识别算法也逐渐表现出了其不足之处。物品识别算法一般用于处理2D图像,而2D图像只能呈现物品在平面上的投影,因此目前的物品识别算法在2D图上很难利用空间距离信息进一步区分处于不同平面的同类物品。以分拣机器人为例,基于单一物品识别算法的分拣机器人[11-13]工作场景的要求较为严苛。如图1所示,其为典型的机械臂分拣作业场景:待分拣的各类物品被无规则地放置在料架上,目标物品识别的难度非常大。由于单一的物品识别算法无法根据场景的空间特征区分目标物品与干扰物,若干扰物堆叠在目标物品周围且出现在分拣机器人摄像头的视野范围内,则会导致物品识别算法错误地将干扰物认定为目标物品,产生错误分拣。
而随着深度摄像机相关技术的成熟与普及,实时获取场景的空间信息已成为可能,深度摄像机也因此被逐渐运用在医学成像[14]、农业采摘[15]、无人驾驶[16]等领域。目前,主流的深度摄像机按技术实现可分为3类:基于结构光技术[17](如Orbbec Gemini 2 L和Orbbec Astra 2)、基于双目视觉技术[18-19](如Intel RealSense D455系列)以及基于飞行时间测量技术[20-21](如微软Azure Kinect)。其中,基于结构光技术的深度摄像头需要发射专门的结构光线(如红外光线)并接收其反射光,形成主动探测。虽然结构光的强度将直接影响测量距离,且这种光线容易受到环境光线以及物品表面对结构光线吸收作用的影响,进一步降低测量精度,但是这种类型的深度摄像机具有低成本与低功耗的优势,能够提供实时且低分辨率的3D点云信息,因此通用性更强。3D点云由深度摄像机捕获的用于表达场景空间立体特征的数据以及由众多包含深度信息的点构成。利用3D点云中的3D空间特征信息弥补单一基于2D图像的物品识别算法在工业分拣场景中的不足,是当前研究的热点。
针对以上问题,当前已有研究者提出将深度学习应用于对3D点云的处理。PointStaClu[22]采用神经网络进行特征提取,并通过稳定聚类判别优化点云聚类任务。还有一些方法采用点云语义分割[23-25]的方式,将具有相似特征的点聚类成同质化的区域。然而,3D点云往往是由上万个点组成,即使在数据预处理阶段,可以以下采样的方式降低数据量,但其规模仍然十分庞大,导致模型在训练和推理阶段造成大量的内存占用与时间消耗。因此基于深度学习的3D点云处理模型很难直接部署到资源受限的嵌入式设备上。
为此我们尝试将深度摄像机与物品识别算法相结合,提出一种基于3D点云探测的轻量化物品识别方法3Deus。3Deus的核心是利用3D点云数据构建一个精确的目标物品及其所处平面的点云模型(简称为目标点云模型)。利用串行聚类集合改进传统的随机样本一致性(random sample consensus,RANSAC)算法,实现目标点云模型的粗略构建;利用平面扩张机制解决了平面点云中平面破碎、边缘损坏以及点云误删的问题,构建更为精细且完整的目标点云模型。3Deus在物品识别层面则通过3D与2D之间的映射,实现3D立体空间信息与2D像素信息之间的一致性绑定;使用通用物品识别算法YOLOv11完成对场景中所有物品的识别,并结合构建的目标点云模型,从所有的识别结果中去除干扰物品,实现对目标物品的充分识别。
由于3Deus在点云探测过程中使用了低复杂度的聚类算法与其他算法实现3D空间特征的提取,且在物品识别算法上可以自由选择算法模型的规模,因此相较于点云深度学习的方法,3Deus更为轻量化,不需要桌面级高性能显卡的支持,硬件成本更低。在实验中,我们也将3Deus分别部署在嵌入式边缘设备Jetson Nano和Jetson NX上,验证了算法实时性,实验结果表明,其更适合物联网边缘设备。
本文的主要贡献包括4个方面:
1) 提出一种基于串行聚类集合改进的RANSAC算法,以解决传统的RANSAC算法在拟合平面过程中无法区分目标物品所在的平面与堆叠物的问题。
2) 为了解决在干扰点云去除过程中出现的目标物品所处平面点云误删的情况,本文提出了平面扩张机制。在删除干扰点云的过程中,进一步判断被删点云所处空间位置,以保留原本属于目标物品所处平面的点云,从而提高平面的完整性。
3) 提出了一种相较于点云深度学习更加轻量化的点云探测与分析方法,具有更低的计算量、更低的硬件成本以及更高的可嵌入性。
4) 为了验证3Deus在区分目标物品与干扰物上的准确性以及在嵌入式设备上的计算效率,我们设计了多种实验场景,分别记录3Deus在不同场景下的测试结果,并且在Jetson Nano与Jetson NX上分别测试算法实时性,3Deus处理时间相比于其他算法,达到了SOTA水平。
随着深度学习的发展,越来越多的研究者将深度学习运用在3D点云处理[26-28]之中。Zeng等人[29]提出GNN-PRFNet,针对现有点云配准拼接未端到端且忽略局部特征的问题,构建含特征提取、配准及拼接融合模块的端到端网络,解决了传统方法局部特征利用不足的问题。Eggen等人[30]提出AdaptNet分类网络和加权激活池化(WAP)模块,设计可动态调整特征权重的池化操作,显著降低了3D点云特征聚合的计算复杂度。Zheng等人[31]提出MATNet网络,针对Transformer忽略点云局部区域及多尺度特征的局限,设计自适应融合多尺度邻域特征的序贯网络,改善了点云拓扑稀疏中多特征聚合不足的问题。Li等人[32]提出深度层次化学习(DHL),针对3D数据缺乏层次化语义解析的问题,构建含层次正则化项、嵌入融合模块的通用框架,结合视觉语言模型自动构建类层次结构,解决了3D任务的语义层次化表征难题。Wu等人[33]针对SNN在3D点云处理的高功耗局限,设计空间感知脉冲神经元(关联点位置与脉冲生成)及3D残差块,实现了在低计算成本之下的高分类精度。
虽然以上神经网络均能实现理想的点云处理效果,但它们需要经过多阶段的训练以及庞大的数据集支撑。且以上神经网络大多采用相邻搜索机制来获取点云间的相邻信息,而这些算法会占用大量的内存资源。因此可以得出,点云深度学习的方式存在一个较为明显的问题,即训练和推理时会消耗大量的时间和计算资源,不适合普及应用,特别是在计算能力受限的边缘设备中。
平面拟合是平面识别中的一个重要处理步骤。为了提高点云平面拟合的精度,已有研究者们提出了多种方法。Holz等人[34]以深度图为基础,通过邻域法计算平面法向量以及平面距离空间坐标原点的距离,进而使用聚类算法得到平面。Poppinga等人[35]在平面拟合中,应用协方差矩阵与均方误差(mean square error,MSE)有效提高了计算效率。Jin等人[36]提出了一种深度平面检测算法,该算法由2部分组成:基于区域增长平面检测和2阶段细化。在第1阶段,通过滑窗法与线性最小二乘法,在深度图中拟合出所有平面,其次计算每个平面的平面度,并按照平面度大小进行排序。在第2阶段,算法根据临近点到平面的距离进行区域面积增长,从而得到最终的平面。EI-Sayed等人[37]首先利用八叉树(oct-tree)将点云划分为多个包含有限点的小立方体,其次根据点云局部密度对各个立方体进行下采样,最终借助主成分分析(PCA)的方式在点云集合中寻找平面。以上算法均能够实现一定的平面拟合效果,但在复杂场景下其鲁棒性较低,难以区分干扰物所处平面与目标物品所处平面,也无法准确地区分目标物品与干扰。
工业场景中,单一2D视觉无法利用空间信息,单一3D点云缺乏纹理特征,因此RGB-D融合识别成为研究热点。但现有方法仍存在“干扰平面区分差、边缘部署难”的问题。基于Faster R-CNN的改进框架[38]通过将点云深度特征融入2D卷积层,采用融合深度信息的改进RoI Pooling模块提升目标定位精度,在分拣场景中识别准确率较高,但模型参数量大,无法在边缘设备高效部署。YOLOv8的RGB-D改进版本[39]通过对称双分支主干网络提取特征,引入多模态特征融合模块实现RGB与深度特征在通道和空间位置的融合,推理时间相对低,但未设计专门的干扰平面去除机制,当地面存在同类物品时,误识别率较高,无法满足分拣机器人的精准抓取需求。ABB机器人的PickMaster5系统[40]结合RGB-D数据与传统模板匹配技术,虽降低了对复杂模型的依赖,但该系统主要适用于规则摆放物品的palletizing场景,在随机堆叠场景中识别率显著下降。
综合上述相关工作的局限性,现有方法无法同时满足轻量化部署、复杂场景抗干扰、平面完整性保障三大核心需求。点云深度学习方法存在计算复杂度高、难部署边缘设备的问题,3Deus通过摒弃复杂神经网络,采用“RANSAC改进+轻量级聚类”,将计算量大大降低;传统RANSAC无法区分堆叠干扰且平面易破碎,3Deus通过串行聚类集合区分堆叠物,并利用平面扩张机制修复破碎平面;RGB-D融合方法存在干扰平面区分差、误识别率高的缺陷,3Deus先通过3D点云去除地面/堆叠干扰平面,再与2D识别框映射,误识别率大大降低;针对边缘部署中推理时间长、资源占用高的问题,3Deus整体算法无大规模参数,满足实时需求。
本节将着重介绍3Deus中点云探测部分,该部分主要由干扰物品所处平面点云去除算法、基于聚类集合改进的RANSAC算法、基于平面扩张机制的改进识别算法组成。该部分的主要目的是去除干扰物所处平面与目标物品所处平面(以下简称为目标台)旁的堆叠物,实现目标台以及目标物品的点云分割。RANSAC平面拟合是点云探测的核心,本节也将分析RANSAC平面拟合的执行流程与不足之处,并给出一套解决方案。
一般场景的3D点云会包含多个平面,传统的平面拟合算法并不能区分期望的平面和干扰的平面。另外,干扰物可能会位于多个平台,但一般情况下干扰物都位于地面上。在本文中,我们只讨论干扰物处于地面以及在目标台旁堆叠的情况。
以图2为例,由可视化的3D点云图观察出,场景中包含了多个平面,如目标台平面、桌腿平面以及地面,其中,桌腿因其点云面积相对较小,在该场景中可以忽略。地面的点云面积最大,因而对RANSAC平面拟合效果的影响也是最大的。然而,干扰物品所处平面(即地面)与目标台在3D空间中是平行的,这就意味着在3D空间平面方程Ax+By+Cz+D=0中,干扰物所处平面与目标台的平面方程在前3个系数上是相同的。另外,目标台点云与干扰物品所处平面点云的数量远远大于3D点云数据中的任意一个平面点云的数量,因此RANSAC算法会优先拟合两者。
根据以上分析,我们首先采用RANSAC算法拟合3D点云中最大平面。在此过程中,我们不需要关注RANSAC算法拟合到了干扰物品所处平面还是目标台,只需关注其获取的平面方程。紧接着,我们提取其平面方程中的A,B,C这3个系数,并设置系数D为0,以此来另外生成1个虚拟平面。假若点云数量为N,由于空间中1个点可以用x,y,z坐标来描述,因此点云集合可以用1个矩阵来存储,这个矩阵大小为N×3,用符号CP表示。记平面系数矩阵PC=(A, B, C),计算每个点距离虚拟平面的投影长度:
$ {\boldsymbol{D}}_{i}=\boldsymbol{C}\boldsymbol{P}\times {\boldsymbol{P}\boldsymbol{C}}^{\mathrm{T}}\;, $
Di表示点到平面的投影,大小为N$ \times $1。在3D空间中,投影长度越小,则说明该点在空间中的位置就越靠近底部。因此我们可采用硬阈值的方式实现底端点云去除。Precd是根据实际场景确定的阈值系数,thresd是计算得出的硬距离阈值。
$ {thres}_{\mathrm{d}}=\left(\max {\boldsymbol{D}}_{i}-\min {\boldsymbol{D}}_{i}\right)\cdot {Prec}_{\mathrm{d}}+\min {\boldsymbol{D}}_{i}。 $
最后将大于thresd的点云保留下来,将小于thresd的点云去除,即可将位于目标台下方的干扰物所处平面点云去除。本节算法的流程见算法1,其中省略了RANSAC算法的实现步骤。
算法1. 干扰物品所处平面点云去除算法。
输入:3D点云矩阵CP,迭代次数N,去除点云的百分比Precd;
输出:剩余点云矩阵CP。
① A, B, C, D=RANSAC(CP);
② Di=Cal_DistanceAll(A, B, C, CP);/*计算所有点 云距虚拟平面的距离*/
③ MaxDi = max(Di);
④ MinDi = min(Di);
⑤ 根据式(2)和Precd计算阈值thresd;
⑥ CP = Remove(CP, Di, thresd);/*从CP中删除小 于thresd的点*/
⑦ return CP。
目前,RANSAC算法在拟合平面中具有较高的准确率,但在复杂环境下仍具有不足之处。根据RANSAC算法的执行流程可知,RANSAC算法在随机选取空间中的3个点后即可计算得到一个平面的方程。但在判断其他点是否位于这个平面的依据则是点到平面的距离,若遇到干扰物与目标台高度平齐的情况,则会使得RANSAC算法无法区分干扰物与目标台,因此RANSAC算法在本质上是计算空间中包含点云数量最多的一个扁平空间。
以图3为例,平面A为传统RANSAC算法拟合出来的桌面平面。理想情况下仅需要获取平面A,但由于干扰物堆叠过高且与平面A相平齐,导致RANSAC算法错误地认为堆叠物与平面A相平齐的那部分点云(记为平面B)也属于平面的一部分。然而实际上,平面A与平面B并不属于同一平面,2个平面近乎垂直。另外,由于受到目标台上物品的遮挡,拟合得到的平面必然是破碎的,在点云图中表现为数个孤立的点云簇。因此如何判断孤立的点云簇属于目标台还是干扰物是本节重点解决的问题。
我们提出一种利用串行聚类集合算法对RANSAC算法进行改进。其中,聚类集合包括了基于密度的聚类DBSCAN(density-based spatial clustering of applications with noise)、K-维树(K-dimensional tree,KD-Tree)、K-均值(K-Means)三种聚类算法。将以上3种聚类算法按照一定的顺序执行,对传统RANSAC算法进行改进,实现干扰点云的滤除与目标点云模型的粗略构建。
首先,利用RANSAC算法拟合剩余点云中的目标台平面,得到平面方程$A_{\rm T}x+B_{\rm T}y+C_{\rm T}z+D_{\rm T}=0 $记平面系数PA=(AT, BT, CT)。其次,借助DBSCAN密度聚类,将拟合出来的平面按照密度划分为多个点云簇,并分别标记。利用KD-Tree计算每一个点云簇内的平均法向量。在代码实现方面,使用Open3d函数库中的函数estimate_normals。用PNa记录所有点云簇的平均法向量,若点云簇数量为Ma,则PNa的大小为Ma×3。计算每个点云簇的平均法向量PNa和目标台平面法向量PA之间的夹角,计算式如式(3)所示。其中|·|运算符表示对向量取模。
$ \boldsymbol{A}\boldsymbol{g}=\mathrm{arccos}\left(\frac{{\boldsymbol{P}\boldsymbol{N}}_{\mathrm{a}}\times {\boldsymbol{P}\boldsymbol{A}}^{\mathrm{T}}}{\left| \boldsymbol{P}{\boldsymbol{N}}_{\mathrm{a}}\right| \cdot \left| \boldsymbol{P}\boldsymbol{A}\right| }\right)\;。$
通过转换将角度限制在[0,90]之间,得到一个大小为Ma×1的向量Ag。最后,借助K-Means对向量Ag进行1D聚类,使其分为正常与异常2类数据,数值较小的为正常类,较大的为异常类,删除异常点云簇后即可获得最终粗略构建的目标点云模型。本节的具体算法流程见算法2。
算法2. 串行聚类集合改进的RANSAC算法。
输入:点云矩阵CP;
输出:正常平面点云簇PNor。
① AT,BT,CT,DT,Plane=RANSAC(CP);/*Plane为 拟合平面的点云矩阵*/
② clusterList=dbscan(Plane);/*clusterList表示所 有的点云簇*/
③ PNa=estimate_normals(clusterList);/*计算所有 点云簇的平均法向量*/
④ 根据式(3)计算Ag,并将Ag限制在[0,90]的 范围内;
⑤ Nor,abNor=Kmeans(Ag);/*借助Kmeans算法 将Ag分为异常和正常角度,Nor一类为正 常,abNor一类为异常*/
⑥ 将所有正常角度对应的点云簇记为PNor;
⑦ return PNor。
凸包能从庞大的点云数据中迅速找到具有一定特征的目标点云,因此凸包计算在3D点云处理的过程中具有非常重要的意义。在3D点云处理的过程中,凸包提取具有一定特征的目标点云时,能够高效确定庞大点云数据的外部边界。通过计算凸包,我们能够得到点云数据的最外层几何形状,这对于后续的目标识别、路径规划、空间优化等应用至关重要。
在3Deus中,2次凸包计算分别基于3D点云和2D图像,前者能够将分离的点云簇包裹成一个统一的平面整体,后者能够根据平面整体进一步获取最外围凸包的点。
首先,针对2.2节中获取的目标台点云计算凸包,将所有分离的点云簇合并为一个平面整体,并记录下所有的3D凸包点。根据坐标转换公式将所有3D凸包点映射到2D图上。转换公式如下:
$ {\boldsymbol{M}}_{\text{camera}}=\left(\begin{matrix}{f}_{x}&0&{pp}_{x}&0\\0&{f}_{y}&{pp}_{y}&0\\0&0&1&0\end{matrix}\right)\;, $
其中Mcamera表示摄像机的内部参数矩阵,fx表示像素宽度方向上的焦距,fy表示像素高度方向上的焦距。ppx与ppy表示主点在像素坐标系下的坐标,这里主点通常表示像素的中心点。不同的深度摄像头具有不同的内部参数矩阵,其内部参数矩阵皆可从指定深度摄像头的相关说明文档中获取。
$ s\left(\begin{matrix}u\\v\\1\end{matrix}\right)={\boldsymbol{M}}_{\text{camera}}\left(\begin{matrix}{x}_{\mathrm{c}}\\{y}_{\mathrm{c}}\\{z}_{\mathrm{c}}\end{matrix}\right)\;, $
$ \left\{\begin{aligned}&su={x}_{\mathrm{c}}\cdot {f}_{x}+{pp}_{x}\cdot {z}_{\mathrm{c}}\;,\\&sv={y}_{\mathrm{c}}\cdot {f}_{y}+{pp}_{y}\cdot {z}_{\mathrm{c}}\;, \\&s={z}_{\mathrm{c}}\;,\end{aligned}\right.\Rightarrow\left\{\begin{aligned}&u=\frac{{x}_{\mathrm{c}}}{{z}_{\mathrm{c}}}\cdot {f}_{x}+{pp}_{x}\;。\\&v=\frac{{y}_{\mathrm{c}}}{{z}_{\mathrm{c}}}\cdot {f}_{y}+{pp}_{y}\;。\end{aligned}\right. $
式(5)与式(6)描述了如何利用内部参数矩阵将3D点云映射到2D图像中。其中,s表示比例系数,xc,yc,zc表示某个点的3D空间坐标,u和v则表示该点在图像上的坐标。完成3D点云到2D图像的映射之后,利用Graham算法计算2D图像的凸包,获取2D图像的目标台轮廓,并记录所有的2D凸包点。
在干扰点云去除和平面扩张机制中需要找出以上2D凸包点在3D点云中的坐标位置。但由于坐标转换过程是不可逆的,即2D坐标不能逆向推导出3D坐标,因此我们采用了坐标映射的方式,通过建立坐标映射表CorTable来保存凸包点在3D点云与2D图像上的坐标映射关系。令凸包点的数量为K,则坐标映射表CorTable的大小为K×5。CorTable的前3列表示凸包点在3D点云中的坐标,后2列表示其在图像上的2D坐标。
根据获取的凸包点以及坐标映射表,将2D图像中相邻的2个凸包点映射回3D点云,并获取其3D坐标,再结合2.2节中利用传统RANSAC算法拟合出的目标台法向量,即可通过计算确定3D空间中经过指定的2个点且平行于一个指定法向量的平面。将点云坐标(X,Y,Z)代入平面方程,计算得到$A_{\rm E} x+ B_{\rm E} y+C_{\rm E} z+D_{\rm E} $的值,并判断其与0之间的大小关系,即可确定对应的点是否为干扰物点云。
然而,由于物品的遮挡而导致的点云破碎性,拟合的凸包并没有完全包裹住目标台,导致算法错误地将原本位于目标台边缘的点云判断为干扰物,使得目标台边缘物品的区分精度下降。为此,3Deus采用了一种平面扩张机制。其步骤如下:
1)输入最小点云阈值minPoint,扩张系数exps。
2)将开始确定的平面向外扩张一定的距离。扩张后的平面方程,除了系数DE外,其他系数与原平面相同。扩张后的平面方程为$A_{\rm E} x+B_{\rm E} y+C_{\rm E} z+D_{\rm plus}= 0 $,Dplus表示为
$ {D}_{\text{plus}}={D}_{\mathrm{E}}\pm \frac{\sqrt{{{{A}_{\mathrm{E}}^{2}}}+{{{B}_{\mathrm{E}}^{2}}}+{{{C}_{\mathrm{E}}^{2}}}}}{exps}。 $
3)检查扩张后与扩张前2个平面之间的点云数量,若大于最小点云阈值minPoint,则说明2个平面内的点属于目标台,将2个平面同时向外移动1个单位的距离,继续执行步骤2)3)。假若小于minPoint,则说明当前已经遍历到目标台以外的部分,则退出当前平面,转向下一个平面继续执行以上的算法,直至遍历完所有的平面。
平面扩张机制解决了在干扰点云去除过程中出现的目标物品所处平面点云误删的情况,在删除干扰点云的过程中,进一步判断被删点云所处空间位置以保留原本属于目标物品所处平面的点云,从而提高平面的完整性。
本节将按照图4所示的算法执行流程,拆解各个步骤并展示其计算结果,以直观且详细地描述3Deus中点云探测的执行逻辑。
通过深度相机获取场景的初始点云与RGB图,如图5所示。根据2.1节中的干扰物所处平面点云去除算法(即算法1),借助RANSAC算法拟合场景中的任意平面并创造一个用于计算投影距离的虚拟平面。根据式(1)计算点云投影距离,设置Precd=0.6,根据式(2)计算点云距离阈值thresd并实现地面点云的去除。
在当前场景中,干扰物所处平面即为地面,因此其效果为地面点云被去除,如图6所示。
根据图6点云信息,再进行一次RANSAC平面拟合,即可获取目标台平面,如图7(a)所示。如图7(b),观察平面点云在整体点云中的位置可知,该算法将堆叠物的点云拟合到了平面中,并且目标台的点云具有极其严重的破碎性,因此需要借助串行聚类集合作进一步处理。
利用DBSCAN算法将平面划分成多个点云簇,该算法实现所给出的案例中总共可划分为4个点云簇,每个点云的法向量可视化如图7(c)所示。可以观察到,异常点云簇的法向量与目标台法向量具有较为明显的差距。通过比较每个点云簇法向量与先前拟合平面的法向量之间的夹角,即可标记并去除异常的点云簇,如图7(d)所示。
利用Open3D中的函数compute_convex_hull对3D点云图进行凸包计算,图8(a)(b)为凸包的可视化结构,将3D凸包点通过坐标转换公式映射到2D图中。
如图8(c)所示,可以观察到在预期目标台中产生了大量的内部干扰点,因而需要借助Graham算法再次计算凸包,获取目标台大致轮廓,如图8(d)所示)。与此同时,生成2D与3D之间的坐标映射表。借助坐标映射表,将2D图中的凸包点映射到3D点云空间中,根据式(7)计算得到目标台边缘轮廓平面。
为了直观展示,这里将计算得到的平面绘制在点云图中,如图9(a)所示。另外,观察图9(b)我们可以发现,凸包并没有完全包裹住目标台,而是位于平面边缘且略微往目标台中心方向收缩的位置,方框内的所有点属于目标点云,若不采用平面扩张机制,则会导致位于目标点云边缘的点被误删,从而导致目标点云缺乏完整性。
为了防止出现点云误删的情况,我们在3Deus中加入了平面扩张机制,通过设定阈值minPoint来判断2个平面之间所包含的点属于目标点云还是属于干扰物。如图9(c)(d)所示,将加入与不加入平面扩张机制的效果进行对比可以发现,平面扩张机制对保留目标台点云的完整性具有重要的意义,图9(d)即为最终分析得出的目标台点云。
另外,将目标台点云的2D图与最初点云的2D图进行对比,如图9(e)(f)所示,3Deus去除了桌外的干扰物,较为精准地保留了目标台点云。将先前去除的地面点云与目标台旁的堆叠干扰点云转化到2D图,如图9(g)(h)所示。最后,借助自训练的YOLOv11模型,对初始的RGB图进行预测,得到每个物品的识别框。将识别框映射到图9(f)(g)(h)中,根据像素信息即可判断物品处于哪一个位置。
3Deus不仅在聚焦的桌上目标物、桌下地面干扰、桌旁堆叠干扰共存的场景中表现出优异性能,还可灵活适配多类具有“物品杂乱摆放、多平面干扰、需精准区分空间维度目标与干扰”特征的复杂场景:在智能仓储中,可用于多层货架的物品识别,精准区分各层货架目标物品与立柱、地面干扰物,适配仓储移动机器人的现场识别需求;在工业装配线中,能识别传送带上歪斜、堆叠的精密零件,排除两侧工具架干扰,满足装配线实时检测诉求;在智能零售自助结算场景中,可区分结算台表面目标商品与手机、购物袋等干扰物品,匹配自助设备低硬件成本部署需求等。这些场景均需“区分空间干扰、精准识别无序物品”,3Deus的应用能力可直接适配,进一步验证其应用普适性,拓展了在物联网感知层物品识别任务中的应用范围。
目标台与摄像头之间的距离在1~1.3 m,其相对位置如图10(a)所示。本实验采用Intel RealSense SR300深度摄像头作为场景的点云探测工具,摄像头外观如图10(b)所示。由于SR300摄像头的缺陷以及实验场地的限制,黑色的地板、透明的水瓶、金属制的易拉罐等物品表面会吸收由SR300发射出来用于检测场景深度的结构光线,因而使得深度摄像头无法接收反射回来的结构光线而丢失某些表面的深度信息。
在本场景中,大面积的黑色地面会吸收由SR300发射出来的结构光而丢失深度信息,因此,为了使得场景更贴近实际应用,我们在地板上放置了大量彩色卡纸,避免地板的深度信息丢失。深度摄像头是本实验的关键,针对当前实验场景,我们仔细调整了摄像头的各个参数,其关键数值如表1所示。另外,为了验证算法在嵌入式设备上的可行性,将3Deus分别部署在了Jetson Nano与Jetson NX上,如图10(b)所示。
在本文提出的算法中,关键参数的设置对系统性能有着显著影响。首先,干扰平面去除阈值Precd设置为0.6。若该值过小,会导致地面干扰点云去除不足;若该值过大,则会过度去除点云,误删目标台有效边缘,导致平面残缺。深度相机参数(如depth accuracy,filter_option等)共同决定了原始点云的质量与密度。对于平面扩张机制,参数minPoint决定了平面扩张的终止条件:其值过小会使平面过度扩张,将邻近干扰物纳入目标平面;其值过大则会导致扩张提前终止,无法得到完整的目标平面轮廓。扩张系数(exps)控制着扩张的步长与精度:过小的exps值会使扩张过程缓慢且对噪声敏感;过大的exps值则会降低边界定位精度,导致平面轮廓扭曲。算法性能依赖于上述参数的协同优化,若调整任一参数,均会使3Deus的整体性能下降。
针对本文所提出的3Deus,我们模拟了一系列实验场景,用于测试3Deus的效果,实验场景由多张白色方桌与圆桌、多张彩色卡纸、大量目标物品以及数个物品识别模型之外的物品组成;目标台上的物品采用常见的日用品,物品摆放形式多样,包括且不限于正放、倒放、斜放、堆叠,物品之间存在倚靠和遮挡,其目的在于使得目标台平面点云具有一定的破碎性,以测试3Deus对于不同场景的适应性。
我们按照之前所描述的实验装置与场景设置,手动拍摄数据集,并根据干扰物的数量、空间位置以及目标台摆放形式,对场景进行分类,具体实验场景如图11和图12所示。
图11中,干扰场景1为仅目标台上有目标物品。干扰场景2为目标台上有目标物品、目标台下有干扰物。干扰场景3为目标台上有目标物品、目标台下有干扰物、目标台旁有堆叠物。干扰场景4在干扰场景3的基础上将目标台斜放。在以上4种干扰场景的目标台上,放置一定数量的目标物品,目标物品皆为深度学习物品识别模型中所涉及的。在目标台下(即地上),也会放置一定数量的物品识别模型内的物品,同时也会在远处放置一定数量与模型无关的物品,但受到深度摄像头的参数限制,远处的物品并不会接收到深度摄像头发射出的结构光,因而不会含有深度信息,也不会对3Deus的效果造成干扰,因此可以忽略不计。数据集的每张图片中,目标物品与干扰物的摆放位置与数量随机。对每种干扰场景拍摄50张图片,同时保存图片对应的点云图,总共拍摄200张图片以及点云图。
同时,为验证3Deus算法的场景可扩展性,我们测试了其在多桌面条件下的性能:
1)如图12(a)所示,设置多目标平面随机布局场景,随机水平摆放1台圆桌与2台方桌,构成3个目标桌面,重点验证算法对多个水平平面以及其上物品的精准识别能力,以及对地面干扰目标的排除能力;
2)如图12(b)所示,为了验证算法对倾斜桌面的识别,增加一个倾斜桌面的方形桌面与正常的水平方形桌面和水平圆形桌面构成组合式实验环境,其中水平方桌和水平圆桌作为目标平面,倾斜方桌作为非目标干扰平面,通过对多桌面的点云提取,需要在识别并剔除倾斜桌面后,仅对水平桌面上的物品进行有效识别,重点验证算法在多目标平面与倾斜干扰平面共存的复杂场景下,对目标平面的区分识别准确性以及稳定性;
3)如图12(c)所示,为了验证算法的场景拓展性,设置多目标不平整平面随机布局场景,随机摆放不平整桌面桌子2台,其上有不平整桌布以模拟不平整平面,在其表面、台体周边空间及台面下方3个空间位置随机摆放物品,重点验证算法在更灵活场景下的可扩展性。
我们对3个实验场景中桌面的摆放位置及摆放物品进行随机调整,并采集数据集。每个实验场景采集200张图像,新增3组实验累计构建600张图像的基准数据集。同时保存图片对应的点云图,进一步扩充数据集维度,全面验证3Deus算法在多水平平面场景、水平加倾斜平面场景以及不平整平面场景中的性能,以确保3Deus在实际应用场景中能有效应对各种干扰以及布局变化。
3Deus的重点在于目标台识别与干扰物去除。在实验中,干扰物包括地面干扰物与堆叠干扰物,为了能够直观地表现模型对目标物与干扰物的区分精度,这里采用了目标检测中常用的指标,即混淆矩阵、准确率(accuracy)、精准率(precision)、召回率(recall)、F1分数。准确率记录了分类正确的样本数占总样本数的比例,精准率记录正确预测为正样本的样本数占预测为正样本总样本数的比例,召回率记录了正确预测为正样本的样本数占实际正样本总样本数的比例。
由于精准率和召回率两者存在一定的权衡关系,即精准度的上升可能会伴随着召回率的下降,因此精准率和召回率并不能完整地对不同方法进行综合衡量,于是,我们引入F1分数以及整体识别准确率,将桌旁与桌上的识别F1分数,用均方根误差(root mean square error,RMSE)计算,得到平面扩张机制对点云探测算法的整体提升效果为30.48%。RMSE的计算式如式(8)所示,其中EU表示桌上物品识别的F1分数差,ED表示桌下物品识别的F1分数差。
$ RMSE=\sqrt{\frac{{EU}^{2}+E{D}^{2}}{2}}\;。 $
另外,本实验还需要测试3Deus在嵌入式设备上的运行时间,因此在统计实验中将记录3Deus在Jetson Nano与Jetson NX上的平均运行时间,并与PC机进行比较。
1)单平面场景统计与对比实验
为了检测3Deus在强干扰场景下的适应性以及关键步骤,同时验证基于聚类集合改进的RANSAC算法和平面扩张机制的重要性,我们引入了2个对比实验。第1个实验为仅包含地面点云去除的算法,在去除地面点云后,直接使用传统RANSAC算法拟合剩余点云中的目标台。第2个实验则是去除平面扩张机制的算法,在点云探测的最后阶段不执行平面扩张机制。具体的对比结果展示在表2~4中。
通过对比实验,我们记录了各实验结果的混淆矩阵。表2展示了3Deus的混淆矩阵,表3展示了仅包含地面点云去除算法的结果,表4则展示了去除平面扩张机制后的算法效果。
根据实验结果,我们可以从表2~4中看到算法在物品识别任务中的表现差异。首先,表2展示了3Deus的混淆矩阵及其识别效果,表明算法在各种场景下都能够较好地识别目标物品,尤其是在桌上和桌下的物品识别中,表现出较高的精准率和召回率。相比之下,表3中的结果显示,在仅去除地面点云的算法中,识别效果有所下降,尤其是在桌旁物品的识别上,误识别和漏识别的情况明显增多。这表明去除地面点云后的传统RANSAC算法在面对复杂的点云数据时,存在一定的局限性。
进一步地,表4展示了去除平面扩张机制后的算法效果。从结果来看,去除平面扩张机制会导致桌旁物品的识别准确率大幅下降,尤其是在复杂场景中,平面扩张机制的缺失导致了识别精度的显著降低。这证明了平面扩张机制在处理边缘物品时的重要性,尤其是在处理物品遮挡或点云破碎性问题时,扩张机制能够有效提升算法的鲁棒性和识别准确性。
综合以上分析,3Deus在所有场景中,对桌上、桌下、桌旁的物品都能达到极高的识别准确率,且关键步骤平面扩张机制和聚类集合改进的RANSAC算法对提高物品识别性能具有重要作用。本文提出的基于聚类集合改进的RANSAC算法和平面扩张机制对于识别准确率有很大的提升。
为验证所提3Deus算法与现有主流算法的性能差异,我们在目标场景中,将3Deus与MO-RANSAC[41],PlaneRCNN[42],MAGSAC[43],PlaneNet[44]等主流算法开展对比实验,以“识别准确率”为核心评价指标,各对比算法的实验结果如下:PlaneNet平均识别准确率为83.47%,PlaneRCNN为87.82%,MAGSAC为88.65%,MO-RANSAC为92.71%,而3Deus的识别准确率达到96.93%,在所有对比算法中实现最优性能;具体来看,3Deus相较于PlaneNet提升13.46个百分点,相较于PlaneRCNN提升9.11个百分点,相较于MAGSAC提升8.28个百分点,即便与性能较优的MO-RANSAC相比,仍实现4.22个百分点的准确率提升,实验结果如图13所示。
3Deus算法实现了最高识别率,主要是因为凭借以下关键技术改进:针对传统算法仅能拟合空间中“点云数量最多的扁平区域”,无法区分目标平面与高度平齐的堆叠干扰物(如桌旁堆叠的同类物品)这一难题,3Deus采用“串行聚类集合改进的RANSAC算法,能够借助点云簇的法向量夹角进行判断,精准滤除堆叠干扰点云,避免将干扰物误判为目标;同时,面对其他算法在物品遮挡场景下易出现的“平面破碎、边缘点云误删”问题(如目标台边缘因物品遮挡导致点云断裂,算法误删边缘目标点云),3Deus通过动态平面扩张机制,可保留原本属于目标平面的边缘点云,修复破碎平面,减少目标物品漏识别。
依据生成的不同场景的数据集,结合评价指标,验证算法的准确性与鲁棒性。根据第2节中对算法的描述,以目标台与干扰物所处平面为基准,将识别目标细分为以下3类:桌上、桌旁和桌下。桌上类属于目标台上需要识别出来的目标物,桌下类和桌旁类属于识别的干扰物。
在3Deus的运行过程中,分别将桌上、桌旁和桌下的点云信息映射到2D图上,如图9(f) (g) (h)所示。经YOLOv11物品识别模型,得到RGB图中每个物品的识别框,并将其映射到点云映射的2D图中,根据识别框内白色像素信息的数量从而判断物品的位置情况。对各种位置情况,分别统计混淆矩阵,计算准确率、精准率、召回率、F1分数。
混淆矩阵是一种评估分类模型性能的表格化方式,但并不能直观地体现模型之间的性能优劣。因此,根据以上统计出的混淆矩阵,我们计算了每一种场景下,桌上、桌下、桌旁物品的识别精准率和召回率,以及整体的识别准确率,将其计算结果绘制在柱状图上。在每一个干扰场景中,将3种对比算法绘制在一起以方便比较,如图14~16所示,可以直观地对比出不同算法在不同干扰场景下对物品识别的水平,反映出探测算法中关键步骤的重要性。图14~16分别表示桌上物品、桌下物品、桌旁物品的识别精准率与召回率。
由图14~16观察得出,3Deus在每一个干扰场景,对所有位置物品的识别效果都保持在极高的水准;对于桌下物品,3种算法都能达到接近100%的精准率,这是因为3种算法都包含了底端点云去除步骤,能够区分出桌下(即地面)的干扰物。因此该算法具有极高的鲁棒性,可适配各类场景并有效去除地面干扰物。地面干扰物的去除在算法实现上较为简单,后续实验结果分析中不再赘述。
与3Deus相比,另外2种算法对于桌上物品和桌旁物品的识别效果有一定程度的下降。去除平面扩张机制的算法对桌上物品识别的召回率相较于另外2个算法有27个百分点左右的下降比例(如图14(b)所示),这是因为破损的平面会导致凸包算法不能完整地包裹住目标台,使得目标台点云的边缘部分被过多地删除,而桌上目标物品的识别需要依靠目标台点云信息,因而造成了在物品识别过程中遗漏了位于目标台边缘位置的物品。或者说,错误地将其识别为桌旁物品,从而导致其召回率下降。但桌下物品与桌旁物品不会被错识别为桌上物品,因此对于桌上物品的识别,3种算法都具有较高的精准率(如图14(a)所示)。相对应的,由图16(a)可以得知,去除平面扩张机制将导致桌上物品大量地被识别为桌旁物品,导致对于桌旁物品识别的精准率急剧下降。然而,由图16(b)可以得知,原本位于桌旁的物品却能准确识别出来。因此,对于桌旁物品识别具有较高召回率。
此外,仅包含地面点云去除的算法,虽然对于桌下物品和桌上物品的识别能达到较高的精准率,但该算法在去除干扰物所处平面点云之后,直接采用传统的RANSAC算法进行平面拟合,目标台两端的堆叠物全部纳入到目标台平面点云中,因此无法对桌旁物品和桌上物品进行区分。观察图14和图16中的干扰场景3和干扰场景4的统计数据可知,3Deus对于桌上物品识别具有较高的召回率和较低的精准率,这是因为原本的桌旁物品几乎全部被识别为桌上物品,而原本位于桌上的物品却能够全部被识别正确。对于桌旁物品,其精准率与召回率下降接近于0,由此也可以体现出传统RANSAC算法在堆叠物与目标台识别上的局限性。
如图17所示。总体而言,3Deus的性能普遍高于另外2种算法,在不同干扰场景下对目标物以及干扰物的区分率以及F1分数均达到了95%以上。为了体现平面扩张机制在整个算法中的重要性,这里将去除平面扩张机制的算法与3Deus进行比较,在桌上物品的识别效果上,3Deus平均提升了15.15个百分点的F1分数,在对桌旁物品的识别效果上提升了40.35个百分点,在整体识别准确率上提升了13.8个百分点。但是由于整体识别准确率中包含了对桌下物品的识别效果,因此其不具备参考价值。
仅包含去除地面点云的算法在干扰场景1和干扰场景2中对桌上和桌下物品都有较高的F1分数和准确率,这是因为干扰场景1和干扰场景2中不包含桌旁物品,使得传统的RANSAC算法能够适应。然而在干扰场景3和干扰场景4中,该算法性能则急剧下降,这也进一步体现了传统的RANSAC算法无法区分桌上物品和桌旁物品。单平面场景识别结果如图18所示。
2)多平面复杂场景统计与对比实验
首先,将原来单一桌面扩展到3个目标桌面。在3个目标桌面、桌体周边及桌体下方3个空间位置随机设置摆放物品。不在桌面上的物品均视为干扰物。通过该实验旨在验证算法在“物品杂乱分布、多平面干扰共存、需精准区分空间维度目标与干扰”的复杂场景下的识别精度与稳定性,多平面复杂场景识别结果如图19所示。
表5展示了3Deus在多水平桌面实验场景中的识别效果。可以看到,在桌上与桌下类别中,算法依旧保持了较高的精准率与召回率。从混淆矩阵的各场景分类结果可以直观发现,尽管场景中存在多个水平平面以及不同高度的物体干扰,算法仍能够有效捕捉目标物品的空间位置特征(桌上、桌下、桌旁),并在大多数情形下完成准确区分。
在4个干扰场景下算法的准确率、召回率和F1分数整体维持在较高水平,其中干扰场景1的性能最为突出,准确率达到100%,F1分数亦接近99%。这一结果主要得益于核心平面面积较大、结构边界清晰,使算法能够充分利用稳定且显著的几何特征,实现对桌上与桌下物品几乎无误判的精确识别。相比之下,干扰场景2、干扰场景3与干扰场景4的部分指标略有下降(但仍基本保持高于90%),如图20所示。这种变化趋势与表5混淆矩阵中的分类结果相吻合。随着场景复杂度不断提升,3个平面之间的空间关系愈发交织,桌旁区域与邻近平面在高度层级和局部几何特征上更易产生重叠,导致算法在这些区域出现一定程度的特征混淆,进而引发少量误判与漏检。尽管如此,整体性能仍处于较高水准,充分说明算法在面对多平面共存、遮挡增强以及物体布局密集的复杂3D场景,依旧具备稳定的识别能力与良好的空间特征适配性。
其次,进一步考虑实际场景中多目标平面与干扰平面交织共存的复杂情况,为全面验证3Deus算法的场景适应性,本文构建“水平加倾斜桌面实验场景”的复合测试场景,以水平圆桌与水平方桌作为双目标平面,搭配倾斜方桌作为非目标干扰平面,重点考察算法在多目标与非水平干扰交织环境下,对目标平面的区分准确率与定位精度及目标物品识别性能的稳定性。
表6展示了3Deus在“水平加倾斜桌面实验场景”的混淆矩阵及其识别效果,根据表格结果我们可以得出算法在这一交织型复杂场景下都能够较好地识别目标物品。
从图21中可以看出,在“水平加倾斜桌面实验场景”的复合测试场景中,算法在4个干扰场景下的准确率、召回率和F1分数整体依旧维持在较高水平,但随着场景复杂度的提升呈现出一定幅度的波动。其中,干扰场景1与干扰场景2的表现最为优异,各项指标几乎均超过98%。
这一结果主要归因于2张目标平面占据了主要空间区域,使整体几何结构清晰、边界分明,算法能够高效区分主目标平面与倾斜干扰平面,从而显著降低特征混淆,特别是在桌上与桌下类别中表现出更强的识别优势。相比之下,干扰场景3与干扰场景4的指标略有下降(但仍保持在91%~97%)。这类指标下降主要源于双核心桌面与周围物体布局更加密集交织,使场景的空间边界更加复杂,局部几何特征更易受到干扰,尤其是桌旁区域在复杂空间结构影响下更容易产生特征混淆,进而导致少量误判与漏检。尽管如此,整体性能仍保持在较高水准,表明3Deus在多平面相互作用、结构复杂度显著提高的条件下仍具备良好的适应性与鲁棒性。
最后,为了验证3Deus对非理想平面的适应能力及场景拓展性,我们设置多目标不平整平面随机布局场景:随机摆放2台覆盖不平整桌布的桌子以模拟表面起伏环境,并同样在其表面、台体周边空间及台面下方3个空间位置随机摆放物品。本实验重点考察算法在平面存在起伏和形变干扰的复杂情形下,能否依然准确提取目标支撑面并有效识别不同空间维度上的目标。
表7展示了3Deus在不平整桌面实验场景中的识别效果。可以看出,尽管桌布的覆盖导致平面几何特征出现不规则波动,算法在“桌上”与“桌下”类别的判定中依然保持了极高的准确性。分析混淆矩阵可知,在干扰场景1和干扰场景2中,针对核心区域物品的误判极少;而在干扰场景3和干扰场景4中,出现的少量混淆主要集中在“桌旁”与“桌上”的边缘交界处,这主要是由于表面的不平整特性在一定程度上干扰了平面区域的扩张与边界界定。尽管如此,整体超过93%的准确率充分表明,算法依旧能够鲁棒地提取出有效的目标支撑面特征,并实现精准的空间区域划分。
在4个干扰场景下算法的准确率、召回率和F1分数整体维持在较高水平,如图22所示,其中干扰场景1与干扰场景2的性能最为突出,各项指标均十分优异。这一结果主要得益于尽管桌面存在表面起伏,但核心平面的整体几何特征没有完全丢失,使算法能够有效克服微小形变的干扰,实现对桌上与桌下物品几乎无误判的精确识别。
相比之下,干扰场景3与干扰场景4的部分指标略有下降(但F1分数仍保持在94%以上)。这主要是因为不平整桌面的边缘几何特征相对模糊,当物品处于桌旁或边缘位置时,褶皱的桌布容易与临近物体产生局部点云粘连,增加了特征分割的难度,进而导致少量误判。尽管如此,整体性能仍处于较高水准,充分说明算法在面对不平整平面与复杂空间布局共存的挑战时,依旧具备稳定的识别能力与优秀的场景可扩展性。
综上所述,以上3个实验结果充分验证了3Deus不仅在单一固定研究场景中具备优异的目标识别性能,更在多水平平面场景、水平加倾斜平面场景以及不平整平面场景下,均能实现精准、稳定的识别效果,充分彰显了算法强大的场景适配性与灵活扩展能力。
3)时间开销评估
将3Deus分别部署在Jetson Nano与Jetson NX上,记录其在每种场景下的运行平均时间,并与Windows操作系统下的PC机进行对比,PC的CPU配置为Intel® CoreTM i7-10510U,GPU配置为NVIDIA GeForce MX350。
从表8可以看出,尽管PC机的运行时间明显短于嵌入式设备,Jetson NX和Jetson Nano的运行时间分别为4.24 s和12.56 s,这表明该算法在嵌入式设备上运行的条件已经具备。虽然Jetson NX性能较强,能够在较短时间内完成任务,但Jetson Nano也能在合理的时间范围内完成计算,特别适合资源受限的应用场景。因此,该算法不仅适用于高性能PC机,也能够在嵌入式设备上高效运行,满足不同硬件环境下的需求,具备在嵌入式设备上运行的条件。
当前的物品识别技术作为物联网应用感知层的重要组成部分,在面对复杂场景时适应能力偏弱。为此本文提出了一种基于点云探测的复杂场景下物品识别方法——3Deus。该方法利用聚类集合改进了传统的RANSAC算法,通过准确地去除干扰物所处平面点云获取目标台点云。在此基础上进一步利用凸包算法、平面扩张机制以及3维空间关系,准确建立目标物品及其所处平面的点云模型,并结合物品识别模型的输出与点云探测的结果,同步实现目标物品高精度识别与干扰物的准确区分。通过对比实验,我们验证了将3Deus应用在多个复杂场景之后,一旦去除干扰物品后,能够有效提高目标物品的识别精度,且该方法适合部署在计算资源受限的嵌入式设备上。
在未来的工作中,继续探索3Deus在实时性方面的进一步优化的可能性,使其能够在更加复杂的场景中实现更高效的物品识别。
邵智超明确研究挑战,提出解决思路,设计并完成实验和撰写论文;邱杰凡提出问题,参与解决方案设计,对论文撰写与实验方案设计进行指导;戴光麟参与方案设计和撰写论文;李乔宏调研文献以及参与算法设计;陈翰墨调研问题背景,对实验、论文写作进行指导;肖克江对整体研究方向和思路进行指导。

参考文献 引证文献
排序方式:
1
Park J, An G, Lee B N, et al. Real-time CNN-based object detection of prohibited items for X-ray security screening[J]. Radiation Physics and Chemistry, 2025, 232: 112681
2
Zu Mengmeng, Jia Jia, Wang Yuan, et al. Research on the application of target detection and recognition technology based on deep learning in industrial automation[C]//Proc of the 2nd Int Conf on Power, Electrical Engineering, Electronics and Control. Piscataway, NJ: IEEE, 2024: 615−620
3
Pei Jiabin, Wu Xiaoming, Liu Xiangzhi, et al. SGD-YOLOv5: A small object detection model for complex industrial environments[C]//Proc of the 41st Int Joint Conf on Neural Networks. Piscataway, NJ: IEEE, 2024: 1−10
4
Li Shiyang, Chen Jing, Peng Weimin, et al. A vehicle detection method based on disparity segmentation[J]. Multimedia Tools and Applications, 2023, 82(13): 19643−19655
5
Wang Yiqing. An analysis of the Viola-Jones face detection algorithm[J]. Image Processing on Line, 2014, 4: 128−148
6
Hassan M M, Hussein H I, Eesa A S, et al. Face recognition based on gabor feature extraction followed by fastica and LDA[J]. Computers, Materials and Continua, 2021, 68(2): 1637−1659
7
Ramu G, Babu S B G T. Image forgery detection for high resolution images using SIFT and RANSAC algorithm[C]//Proc of the 2nd Int Conf on Communication and Electronics Systems. Piscataway, NJ: IEEE, 2017: 850−854
8
Ravi N, Gabeur V, Hu Yuanting, et al. SAM2: Segment anything in images and videos[J]. arXiv preprint, arXiv: 2408.00714, 2024
9
Khanam R, Hussain M. YOLOv11: An overview of the key architectural enhancements[J]. arXiv preprint, arXiv: 2410.17725, 2024
10
Liu Honyi, Tseng Y, Chang K, et al. A denoising FPN with transformer R-CNN for tiny object detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 1−15
11
Zhu Ran, Chen Silu, Li Xiaocong, et al. Recognition and localization method for high-speed robotic sorting in complex environments based on image interference assessment[J]. International Journal of Intelligent Robotics and Applications, 2025, 9(3): 1173−1186
12
Sidehabi S W, Azis M F, Asbar M. Development of a color-based image recognition system for robotic sorting and picking[J]. INTEK: Jurnal Penelitian, 2024, 11(2): 77−82
13
Titu M F S, Haque S M R, Islam R, et al. Experiments with cooperative robots that can detect object’s shape, color and size to perform tasks in industrial workplaces[J]. International Journal of Intelligent Robotics and Applications, 2024, 8(1): 179−192
14
Ali S, Pandey A K. ArthroNet: A monocular depth estimation technique with 3D segmented maps for knee arthroscopy[J]. Intelligent Medicine, 2023, 3(2): 129−138
15
Liu Sixing, Liu Mingmiao, Chai Yan, et al. Recognition and location of pepper picking based on improved YOLOv5s and depth camera[J]. Applied Engineering in Agriculture, 2023, 39(2): 179−185
16
Thanh P T H, Bui M Q V, Nguyen D D, et al. Transfer multi-source knowledge via scale-aware online domain adaptation in depth estimation for autonomous driving[J]. Image and Vision Computing, 2024, 141: 104871
17
Yang Tao, Gu Feifei. Overview of modulation techniques for spatially structured-light 3D imaging[J]. Optics & Laser Technology, 2024, 169: 110037
18
Zheng Yuanzhou, Liu Peng, Qian Long, et al. Recognition and depth estimation of ships based on binocular stereo vision[J]. Journal of Marine Science and Engineering, 2022, 10(8): 1153
19
Zhang Hongxin, Lee S. Advancing the robotic vision revolution: Development and evaluation of a bionic binocular system for enhanced robotic vision[J]. Biomimetics, 2024, 9(6): 371
20
Bamji C, Godbaz J, Oh M, et al. A review of indirect time-of-flight technologies[J]. IEEE Transactions on Electron Devices, 2022, 69(6): 2779−2793
21
Paredes A L, Song Qiang, Conde M H. Performance evaluation of state-of-the-art high-resolution time-of-flight cameras[J]. IEEE Sensors Journal, 2023, 23(12): 13711−13727
22
Cao Xin, Wang Haoyu, Zhu Qiuquan, et al. PointStaClu: A deep point cloud clustering method based on stable cluster discrimination[J]. Remote Sensing, 2024, 16(13): 2423
23
Wang Jingyi, Liu Yu, Tan Hanlin, et al. A survey on weakly supervised 3D point cloud semantic segmentation[J]. IET Computer Vision, 2024, 18(3): 329−342
24
Noichl F, Collins F C, Braun A, et al. Enhancing point cloud semantic segmentation in the data-scarce domain of industrial plants through synthetic data[J]. Computer-Aided Civil and Infrastructure Engineering, 2024, 39(10): 1530−1549
25
Singh D P, Yadav M. Deep learning-based semantic segmentation of three-dimensional point cloud: A comprehensive review[J]. International Journal of Remote Sensing, 2024, 45(2): 532−586
26
韩冰, 张鑫云, 任爽. 基于三维点云的卷积运算综述[J]. 计算机研究与发展, 2023, 60(4): 873−902
Han Bing, Zhang Xinyun, Ren Shuang. Survey of convolution operations based on 3D point clouds[J]. Journal of Computer Research and Development, 2023, 60(4): 873−902 (in Chinese)
27
吕卓, 姚治成, 贾玉祥, 等. 3D物体检测的异构方法[J]. 计算机研究与发展, 2021, 58(12): 2748−2759
Lü Zhuo, Yao Zhicheng, Jia Yuxiang, et al. A heterogeneous approach for 3D object detection[J]. Journal of Computer Research and Development, 2021, 58(12): 2748−2759 (in Chinese)
28
廖瑞杰, 杨绍发, 孟文霞, 等. SegGraph: 室外场景三维点云闭环检测算法[J]. 计算机研究与发展, 2019, 56(2): 338−348
Liao Ruijie, Yang Shaofa, Meng Wenxia, et al. SegGraph: An algorithm for loop-closure detection in outdoor scenes using 3D point clouds[J]. Journal of Computer Research and Development, 2019, 56(2): 338−348 (in Chinese)
29
Zeng Wenhao, Su Gongbing, Su Zixuan, et al. Research on point cloud registration and stitching fusion algorithm based on GCN-PRFNet[J]. IEEE Access, 2025, 13: 43384−43397
30
Eggen K, Fan Hongchao. WAPooling: An adaptive plug-and-play module for feature aggregation in point cloud classification networks[J]. International Journal of Applied Earth Observation and Geoinformation, 2025, 138: 104439
31
Zheng Yufei, Lu Jian, Chen Xiaogai, et al. MATNet: Semantic segmentation of 3D point clouds with multiscale adaptive transformer[J]. Computers and Electrical Engineering, 2024, 119: 109526
32
Li Chongshou, Liu Yuheng, Li Xinke, et al. Deep hierarchical learning for 3D semantic segmentation[J]. International Journal of Computer Vision, 2025, 133(7): 4420−4441
33
Wu Qiaoyun, Zhang Quanxiao, Tan Chunyu, et al. Point-to-spike residual learning for energy-efficient 3D point cloud classification[C]//Proc of the 38th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2024, 6092−6099
34
Holz D, Holzer S, Rusu R B, et al. Real-time plane segmentation using RGB-D cameras[C]//Proc of the 15th Robot Soccer World Cup. Berlin: Springer, 2011: 306−317
35
Poppinga J, Vaskevicius N, Birk A, et al. Fast plane detection and polygonalization in noisy 3D range images[C]//Proc of the 21st IEEE/RSJ Int Conf on Intelligent Robots and Systems. Piscataway, NJ: IEEE, 2008: 3378−3383
36
Jin Zhi, Tillo T, Zou Wenbin, et al. Robust plane detection using depth information from a consumer depth camera[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2017, 29(2): 447−460
37
El-Sayed E, Abdel-Kader R F, Nashaat H, et al. Plane detection in 3D point cloud using octree-balanced density down-sampling and iterative adaptive plane extraction[J]. IET Image Processing, 2018, 12(9): 1595−1605
38
Mouzenidis P, Louros A, Konstantinidis D, et al. Multi-modal variational faster R-CNN for improved visual object detection in manufacturing[C]//Proc of the 18th IEEE/CVF Int Conf on Computer Vision. Piscataway, NJ: IEEE, 2021: 2587−2594
39
Song Guozhu, Wang Jian, Ma Rongting, et al. Study on the fusion of improved YOLOv8 and depth camera for bunch tomato stem picking point recognition and localization[J]. Frontiers in Plant Science, 2024, 15: 1447855
40
ABB Robotics. PickMaster twin technical specification[EB/OL]. [2026-01-26]. https://www.abb.com/global/en/areas/robotics/products/software/pickmaster-twin
41
Lim S, Yoo Y, Lee J K, et al. Multi-object RANSAC: Efficient plane clustering method in a clutter[C]//Proc of the 41st IEEE Int Conf on Robotics and Automation. Piscataway, NJ: IEEE, 2024: 3079−3085
42
Liu Chen, Kim K, Gu Jinwei, et al. PlaneRCNN: 3D plane detection and reconstruction from a single image[C]//Proc of the 30th IEEE/CVF Conf on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2019: 4450−4459
43
Barath D, Matas J, Noskova J. MAGSAC: Marginalizing sample consensus[C]//Proc of the 30th IEEE/CVF Conf on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2019: 10197−10205
44
Liu Chen, Yang Jimei, Ceylan D, et al. PlaneNet: Piece-wise planar reconstruction from a single rgb image[C]//Proc of the 31st IEEE Conf on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2018: 2579−2588
2026年第63卷第8期
83
40
0
引用本文
BibTeX
文章信息
doi: 10.7544/issn1000-1239.202550696
  • 接收时间:2025-09-21
  • 首发时间:2026-11-17
  • 出版时间:2026-08-01
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-09-21
  • 修回日期:2026-04-24
  • 录用日期:2026-04-30
基金
作者信息
    1浙江工业大学计算机科学与技术学院 杭州 310023
    2华中师范大学人工智能教育学部 武汉 430079
    3浙江机电职业技术大学 杭州 310053

通讯作者:

戴光麟()
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/10.7544/issn1000-1239.202550696
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
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
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