Article(id=1312006326264365397, tenantId=1146029695717560320, journalId=1311385739772133495, issueId=1311999647825101743, articleNumber=null, orderNo=null, doi=10.6041/j.issn.1000-1298.2025.12.051, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1723392000000, receivedDateStr=2024-08-12, revisedDate=1728835200000, revisedDateStr=2024-10-14, acceptedDate=null, acceptedDateStr=null, onlineDate=1790736284080, onlineDateStr=2026-09-30, pubDate=1766592000000, pubDateStr=2025-12-25, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1790736284080, onlineIssueDateStr=2026-09-30, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1790736284080, creator=13701087609, updateTime=1790736284080, updator=13701087609, issue=Issue{id=1311999647825101743, tenantId=1146029695717560320, journalId=1311385739772133495, year='2025', volume='56', issue='12', pageStart='1', pageEnd='818', issueExtLink='null', onlineDate='null', pubDate='1766592000000', pubDateStr='2025-12-25', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1790734691815, creator='13701087609', updateTime=1790738411040, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1312015247465083866, tenantId=1146029695717560320, journalId=1311385739772133495, issueId=1311999647825101743, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1312015247465083867, tenantId=1146029695717560320, journalId=1311385739772133495, issueId=1311999647825101743, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=560, endPage=567, ext={EN=ArticleExt(id=1312006326482469206, articleId=1312006326264365397, tenantId=1146029695717560320, journalId=1311385739772133495, language=EN, title=Rice Seedling Counting Method Based on Multi-column Convolutional Network Integrating Multi-scale Features, columnId=1311999800569066336, journalTitle=Transactions of the Chinese Society for Agricultural Machinery, columnName=Agricultural Informatization Engineering, runingTitle=null, highlight=null, articleAbstract=

The planting density of rice in the early seedling stage is closely related to yield estimation and field management. With the development of computer vision technology, the use of drone image recognition technology to automatically count rice is gradually replacing traditional manual sampling and counting methods. However, images taken by drones have problems such as irregular rice seedling shapes, overlapping rice seedlings, scale changes, and background occlusion, which lead to deviations in feature learning, which in turn affects the accurate estimation of rice density and the efficiency of computational prediction. Significant relative size differences among rice seedlings will also become another important factor affecting counting accuracy. In response to these problems, a rice seedling counting algorithm (rice counting network, RCN) was proposed based on an improved multi-column convolutional neural network. Based on the size and shape of the target, an improved Gaussian density kernel was proposed, which combined the similarity information of the rice image and the RGB image of the center point to make the density kernel more suitable for the texture of the rice seedlings. The number of rice ears of different sizes and shapes was quantified, and the scales were divided finely to calculate the receptive field sizes of different features that needed to be fused. The fusion expansion convolution module was used to increase the receptive field size and enhance the model's multi-scale capabilities. On this basis, appropriate attention mechanisms were added to the pathways at different scales to enhance attention to reasonable information about rice seedlings at different scales. It showed obvious effectiveness in scenes with complex backgrounds, multi-scales and dense occlusions. On the RSD2023 data set captured at a height of 5 m, RCN's mean absolute error MAE=3.10 holes/m2 and mean square error MSE=4.03 holes2/m4, which were better than that of the mainstream algorithms such as MCNN, CSRNet, SCAR and TasselNetV2.

, authors=Yi PAN1, Haoran SU1, Linlin QIN1, Gang WU1, Chun SHI1, Yikun WANG2, Sheng HUA2, authorsList=Yi PAN, Haoran SU, Linlin QIN, Gang WU, Chun SHI, Yikun WANG, Sheng HUA, authorCompany=null, correspAuthors=Linlin QIN, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, fund=null), CN=ArticleExt(id=1312006329342984565, articleId=1312006326264365397, tenantId=1146029695717560320, journalId=1311385739772133495, language=CN, title=融合多尺度特征与多列卷积网络的水稻苗无人机遥感计数方法, columnId=1311999800749421410, journalTitle=农业机械学报, columnName=农业信息化工程, runingTitle=null, highlight=null, articleAbstract=

水稻早苗期种植密度与产量估算、田间管理密切相关。随着计算机视觉技术的发展,利用无人机图像识别技术自动统计水稻数量正逐步取代传统的人工采样和计数方法。然而,无人机拍摄的图像存在水稻苗形状不规则、水稻苗重叠、尺度变化和背景遮挡等问题,导致特征学习出现偏差,进而影响了水稻密度的准确估算和计算预测的效率。水稻苗之间显著的相对大小差异也成为影响计数精度的另一重要因素。针对这些问题,提出了一种基于改进的多列卷积神经网络的水稻苗计数算法(Rice counting network,RCN)。该算法根据目标的大小和形状,提出了改进高斯密度核,结合水稻图像与中心点RGB图像的相似度信息,使密度核更加贴合水稻苗纹理。对不同大小和形状的稻穗数量进行量化,并对尺度进行精细划分,计算出需要融合的不同特征感受野大小,采用融合膨胀卷积模块,增加了感受野大小,增强了模型对多尺度环境中的处理效果。在此基础上,对不同尺度的通路添加合适的注意力机制,增强不同尺度下对水稻苗合理信息的关注,在复杂背景、多尺度和密集遮挡的场景中体现出了明显的有效性。在高度5 m捕获的RSD2023数据集上,RCN的平均绝对误差为3.10穴/m2,均方误差为4.03穴2/m4,优于MCNN、CSRNet、SCAR和TasselNetV2等主流算法。

, authors=潘义1, 苏浩然1, 秦琳琳1, 吴刚1, 石春1, 王宜坤2, 华胜2, authorsList=潘义, 苏浩然, 秦琳琳, 吴刚, 石春, 王宜坤, 华胜, authorCompany=null, correspAuthors=秦琳琳, authorNote=

潘义(2000—),男,硕士生,主要从事农业信息化技术研究,E-mail:

, correspAuthorsNote=
秦琳琳(1975—),女,高级工程师,博士,主要从事人工环境建模与控制研究,E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=HFwi0NisIav2fqrxBY2h7A==, magXml=ytpiGdRGiaS0nzoEEZuTPQ==, pdfUrl=null, pdf=SA6W9LGPqkL3frH2GQCUNw==, pdfFileSize=6333118, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=sXgpXe+lroTyr1hvKIG4Rg==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=NnK7KVbgSF1JywhyqJXCXQ==, mapNumber=null, fund=null)}, authors=[Author(id=1312006329837912445, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=alexpanyi@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1312006329909215615, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006329837912445, language=EN, stringName=Yi PAN, firstName=Yi, middleName=null, lastName=PAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006329972130176, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006329837912445, language=CN, stringName=潘义, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio={"content":"

潘义(2000—),男,硕士生,主要从事农业信息化技术研究,E-mail:

"}, bioImg=null, bioContent=

潘义(2000—),男,硕士生,主要从事农业信息化技术研究,E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006330047627650, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, 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=1312006330127319428, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330047627650, language=EN, stringName=Haoran SU, firstName=Haoran, middleName=null, lastName=SU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006330215399813, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330047627650, language=CN, stringName=苏浩然, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006330303480199, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=qinll@ustc.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1312006330404143497, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330303480199, language=EN, stringName=Linlin QIN, firstName=Linlin, middleName=null, lastName=QIN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006330475446666, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330303480199, language=CN, stringName=秦琳琳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006330571915660, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, 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=1312006330664190350, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330571915660, language=EN, stringName=Gang WU, firstName=Gang, middleName=null, lastName=WU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006330731299215, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330571915660, language=CN, stringName=吴刚, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006330806796689, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, 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=1312006330899071379, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330806796689, language=EN, stringName=Chun SHI, firstName=Chun, middleName=null, lastName=SHI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006330987151764, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330806796689, language=CN, stringName=石春, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006331071037846, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, 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=1312006331150729624, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006331071037846, language=EN, stringName=Yikun WANG, firstName=Yikun, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006331213644185, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006331071037846, language=CN, stringName=王宜坤, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.安徽喜洋洋农业科技有限公司,合肥 231521, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329678528889, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=2., ext=[AuthorCompanyExt(id=1312006329686917498, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China), AuthorCompanyExt(id=1312006329762414971, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.安徽喜洋洋农业科技有限公司,合肥 231521)])]), Author(id=1312006331280753051, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, orderNo=6, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1312006331352056221, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006331280753051, language=EN, stringName=Sheng HUA, firstName=Sheng, middleName=null, lastName=HUA, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006331456913822, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006331280753051, language=CN, stringName=华胜, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.安徽喜洋洋农业科技有限公司,合肥 231521, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329678528889, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=2., ext=[AuthorCompanyExt(id=1312006329686917498, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China), AuthorCompanyExt(id=1312006329762414971, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.安徽喜洋洋农业科技有限公司,合肥 231521)])])], keywords=[Keyword(id=1312006331574354335, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, orderNo=1, keyword=rice seedlings), Keyword(id=1312006331658240416, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, orderNo=2, keyword=counting), Keyword(id=1312006331737932193, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, orderNo=3, keyword=density map), Keyword(id=1312006331842789794, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, orderNo=4, keyword=multi-column neural network), Keyword(id=1312006331930870179, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, orderNo=5, keyword=UAV remote sensing), Keyword(id=1312006332023144868, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, orderNo=6, keyword=multi-scale feature), Keyword(id=1312006332098642341, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, orderNo=1, keyword=水稻苗), Keyword(id=1312006332165751206, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, orderNo=2, keyword=计数), Keyword(id=1312006332232860071, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, orderNo=3, keyword=密度图), Keyword(id=1312006332287386024, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, orderNo=4, keyword=多列神经网络), Keyword(id=1312006332371272105, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, orderNo=5, keyword=无人机遥感), Keyword(id=1312006332455158186, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, orderNo=6, keyword=多尺度特征)], refs=[Reference(id=1312006335147901386, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=国家统计局, journalName=中国统计年鉴2023, refType=null, unstructuredReference=国家统计局. 中国统计年鉴2023[M]. 北京: 中国统计出版社, 2023., articleTitle=null, refAbstract=null), Reference(id=1312006335227593163, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=梁青铎, journalName=null, refType=null, unstructuredReference=梁青铎. 施氮量和机插密度对水稻生长、产量和氮肥利用效率的影响[D]. 北京: 中国农业科学院, 2020., articleTitle=施氮量和机插密度对水稻生长、产量和氮肥利用效率的影响, refAbstract=null), Reference(id=1312006335311479244, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=LIANG Qingduo, journalName=null, refType=null, unstructuredReference=LIANG Qingduo. Effects of nitrogen application rate and planting density on growth, yield and nitrogen use efficiency of rice[D]. Beijing: Chinese Academy of Agricultural Sciences, 2020. (in Chinese), articleTitle=Effects of nitrogen application rate and planting density on growth, yield and nitrogen use efficiency of rice, refAbstract=null), Reference(id=1312006335391171021, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2021, volume=4, issue=4, pageStart=20213, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=ALIPOUR F, MOBASSER H R, journalName=Agrosystems, Geosciences & Environment, refType=null, unstructuredReference=ALIPOUR F, MOBASSER H R. Effect of planting density on growth characteristics and grain yield increase in successive cultivations of two rice cultivars[J]. Agrosystems, Geosciences & Environment, 2021, 4(4): 20213., articleTitle=Effect of planting density on growth characteristics and grain yield increase in successive cultivations of two rice cultivars, refAbstract=null), Reference(id=1312006335462474190, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2023, volume=76, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[4], rfOrder=4, authorNames=PICKSON R, GUI P, CHEN A, journalName=Ecological Informatics, refType=null, unstructuredReference=PICKSON R, GUI P, CHEN A, et al. Climate change and food security nexus in Asia: a regional comparison[J]. Ecological Informatics, 2023, 76: 102038., articleTitle=Climate change and food security nexus in Asia: a regional comparison, refAbstract=null), Reference(id=1312006335529583055, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2022, volume=12, issue=3, pageStart=1047, pageEnd=null, url=null, language=null, rfNumber=[5], rfOrder=5, authorNames=ASLAN M, DURDU A, SABANCI K, journalName=Applied Sciences, refType=null, unstructuredReference=ASLAN M, DURDU A, SABANCI K, et al. A comprehensive survey of the recent studies with UAV for precision agriculture in open fields and greenhouses[J]. Applied Sciences, 2022, 12(3): 1047., articleTitle=A comprehensive survey of the recent studies with UAV for precision agriculture in open fields and greenhouses, refAbstract=null), Reference(id=1312006335621857744, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2012, volume=13, issue=null, pageStart=693, pageEnd=712, url=null, language=null, rfNumber=[6], rfOrder=6, authorNames=ZHANG C, KOVACS J M, journalName=Precision Agriculture, refType=null, unstructuredReference=ZHANG C, KOVACS J M. The application of small unmanned aerial systems for precision agriculture: a review[J]. Precision Agriculture, 2012, 13: 693-712., articleTitle=The application of small unmanned aerial systems for precision agriculture: a review, refAbstract=null), Reference(id=1312006335697355217, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2017, volume=60, issue=6, pageStart=84, pageEnd=90, url=null, language=null, rfNumber=[7], rfOrder=7, authorNames=KRIZHEVSKY A, SUTSKEVER I, HINTON G E, journalName=Communications of the ACM, refType=null, unstructuredReference=KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90., articleTitle=ImageNet classification with deep convolutional neural networks, refAbstract=null), Reference(id=1312006335764464082, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=8, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=[8], rfOrder=8, authorNames=KITANO B T, MENDES C C, GEUS A R, journalName=IEEE Geoscience and Remote Sensing Letters, refType=null, unstructuredReference=KITANO B T, MENDES C C, GEUS A R, et al. Corn plant counting using deep learning and UAV images[J]. IEEE Geoscience and Remote Sensing Letters, 2019(8): 1-5., articleTitle=Corn plant counting using deep learning and UAV images, refAbstract=null), Reference(id=1312006335831572947, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2023, volume=50, issue=7, pageStart=74, pageEnd=82, url=null, language=null, rfNumber=[9], rfOrder=9, authorNames=王宏乐, 叶全洲, 王兴林, journalName=广东农业科学, refType=null, unstructuredReference=王宏乐, 叶全洲, 王兴林, 等. 基于YOLOv7的无人机影像稻穗计数方法研究[J]. 广东农业科学, 2023, 50(7): 74-82., articleTitle=基于YOLOv7的无人机影像稻穗计数方法研究, refAbstract=null), Reference(id=1312006335923847636, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2023, volume=50, issue=7, pageStart=74, pageEnd=82, url=null, language=null, rfNumber=[9], rfOrder=10, authorNames=WANG Hongle, YE Quanzhou, WANG Xinglin, journalName=Guangdong Agricultural Sciences, refType=null, unstructuredReference=WANG Hongle, YE Quanzhou, WANG Xinglin, et al. Rice panicles counting method based on YOLOv7 using unmanned aerial vehicles images[J]. Guangdong Agricultural Sciences, 2023, 50(7): 74-82. (in Chinese), articleTitle=Rice panicles counting method based on YOLOv7 using unmanned aerial vehicles images, refAbstract=null), Reference(id=1312006336003539413, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2008, volume=19, issue=5, pageStart=617, pageEnd=627, url=null, language=null, rfNumber=[10], rfOrder=11, authorNames=HANBURY A, journalName=Journal of Visual Languages & Computing, refType=null, unstructuredReference=HANBURY A. A survey of methods for image annotation[J]. Journal of Visual Languages & Computing, 2008, 19(5): 617-627., articleTitle=A survey of methods for image annotation, refAbstract=null), Reference(id=1312006336100008406, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2021, volume=12, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[11], rfOrder=12, authorNames=WANG Y, QIN Y, CUI J, journalName=Frontiers in Plant Science, refType=null, unstructuredReference=WANG Y, QIN Y, CUI J, et al. Occlusion robust wheat ear counting algorithm based on deep learning[J]. Frontiers in Plant Science, 2021, 12: 645899., articleTitle=Occlusion robust wheat ear counting algorithm based on deep learning, refAbstract=null), Reference(id=1312006336179700183, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2010, volume=1, issue=null, pageStart=1324, pageEnd=1332, url=null, language=null, rfNumber=[12], rfOrder=13, authorNames=LEMPITSKY V, ZISSERMAN A, journalName=Advances in Neural Information Processing Systems (NIPS), refType=null, unstructuredReference=LEMPITSKY V, ZISSERMAN A. Learning to count objects in images[J]. Advances in Neural Information Processing Systems (NIPS), 2010, 1: 1324-1332., articleTitle=Learning to count objects in images, refAbstract=null), Reference(id=1312006336242614744, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=589, pageEnd=597, url=null, language=null, rfNumber=[13], rfOrder=14, authorNames=ZHANG Y, ZHOU D, CHEN S, journalName=null, refType=null, unstructuredReference=ZHANG Y, ZHOU D, CHEN S, et al. Single-image crowd counting via multi-column convolutional neural network[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, 2016: 589-597., articleTitle=Single-image crowd counting via multi-column convolutional neural network, refAbstract=null), Reference(id=1312006336326500825, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=1091, pageEnd=1100, url=null, language=null, rfNumber=[14], rfOrder=15, authorNames=LI Y, ZHANG X, CHEN D, journalName=null, refType=null, unstructuredReference=LI Y, ZHANG X, CHEN D, et al. CSRNet: dilated convolutional neural networks for understanding the highly congested scenes[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, 2018: 1091-1100., articleTitle=CSRNet: dilated convolutional neural networks for understanding the highly congested scenes, refAbstract=null), Reference(id=1312006336397803994, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2020, volume=11, issue=null, pageStart=617, pageEnd=null, url=null, language=null, rfNumber=[15], rfOrder=16, authorNames=ZHANG J, ZHAO B, YANG C, journalName=Frontiers in Plant Science, refType=null, unstructuredReference=ZHANG J, ZHAO B, YANG C, et al. Rapeseed stand count estimation at leaf development stages with UAV imagery and convolutional neural networks[J]. Frontiers in Plant Science, 2020, 11: 617., articleTitle=Rapeseed stand count estimation at leaf development stages with UAV imagery and convolutional neural networks, refAbstract=null), Reference(id=1312006336481690075, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2019, volume=11, issue=6, pageStart=691, pageEnd=null, url=null, language=null, rfNumber=[16], rfOrder=17, authorNames=WU J, YANG G, YANG X, journalName=Remote Sensing, refType=null, unstructuredReference=WU J, YANG G, YANG X, et al. Automatic counting of in situ rice seedlings from UAV images based on a deep fully convolutional neural network[J]. Remote Sensing, 2019, 11(6): 691., articleTitle=Automatic counting of in situ rice seedlings from UAV images based on a deep fully convolutional neural network, refAbstract=null), Reference(id=1312006336552993244, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2019, volume=15, issue=1, pageStart=150, pageEnd=null, url=null, language=null, rfNumber=[17], rfOrder=18, authorNames=XIONG H, CAO Z, LU H, journalName=Plant Methods, refType=null, unstructuredReference=XIONG H, CAO Z, LU H, et al. TasselNetv2: in-field counting of wheat spikes with context-augmented local regression networks[J]. Plant Methods, 2019, 15(1): 150., articleTitle=TasselNetv2: in-field counting of wheat spikes with context-augmented local regression networks, refAbstract=null), Reference(id=1312006336620102109, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2023, volume=54, issue=3, pageStart=266, pageEnd=273, url=null, language=null, rfNumber=[18], rfOrder=19, authorNames=饶雄飞, 周龙宇, 杨春雷, journalName=农业机械学报, refType=null, unstructuredReference=饶雄飞, 周龙宇, 杨春雷, 等. 基于无人机多光谱影像和关键点检测的雪茄烟株数提取[J]. 农业机械学报, 2023, 54(3): 266-273., articleTitle=基于无人机多光谱影像和关键点检测的雪茄烟株数提取, refAbstract=null), Reference(id=1312006336691405278, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2023, volume=54, issue=3, pageStart=266, pageEnd=273, url=null, language=null, rfNumber=[18], rfOrder=20, authorNames=RAO Xiongfei, ZHOU Longyu, YANG Chunlei, journalName=Transactions of the Chinese Society for Agricultural Machinery, refType=null, unstructuredReference=RAO Xiongfei, ZHOU Longyu, YANG Chunlei, et al. Counting cigar tobacco plants from UAV multispectral images via key points detection approach[J], Transactions of the Chinese Society for Agricultural Machinery, 2023, 54(3): 266-273. (in Chinese), articleTitle=Counting cigar tobacco plants from UAV multispectral images via key points detection approach, refAbstract=null), Reference(id=1312006336766902751, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2021, volume=23, issue=6, pageStart=721, pageEnd=null, url=null, language=null, rfNumber=[19], rfOrder=21, authorNames=FENG A, LI H, LIU Z, journalName=Entropy, refType=null, unstructuredReference=FENG A, LI H, LIU Z, et al. Research on a rice counting algorithm based on an improved MCNN and a density map[J]. Entropy, 2021, 23(6): 721., articleTitle=Research on a rice counting algorithm based on an improved MCNN and a density map, refAbstract=null), Reference(id=1312006336842400224, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2024, volume=116, issue=3, pageStart=810, pageEnd=825, url=null, language=null, rfNumber=[20], rfOrder=22, authorNames=YANG H, YUAN H, WANG Y, journalName=Agronomy Journal, refType=null, unstructuredReference=YANG H, YUAN H, WANG Y, et al. Rape seedling density estimation in-field conditions based on improved multi-column convolutional neural network[J]. Agronomy Journal, 2024, 116(3): 810-825., articleTitle=Rape seedling density estimation in-field conditions based on improved multi-column convolutional neural network, refAbstract=null), Reference(id=1312006336930480609, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2023, volume=211, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[21], rfOrder=23, authorNames=CHEN Y, JIANG H, ZHANG X, journalName=Computers and Electronics in Agriculture, refType=null, unstructuredReference=CHEN Y, JIANG H, ZHANG X, et al. Refined feature fusion for in-field high-density and multi-scale rice panicle counting in UAV images[J]. Computers and Electronics in Agriculture, 2023, 211: 108032., articleTitle=Refined feature fusion for in-field high-density and multi-scale rice panicle counting in UAV images, refAbstract=null), Reference(id=1312006336993395170, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=10397, pageEnd=10406, url=null, language=null, rfNumber=[22], rfOrder=24, authorNames=LAN S, REN Z, WU Y, journalName=null, refType=null, unstructuredReference=LAN S, REN Z, WU Y, et al. SaccadeNet: a fast and accurate object detector[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, 2020: 10397-10406., articleTitle=SaccadeNet: a fast and accurate object detector, refAbstract=null), Reference(id=1312006337064698339, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=3, pageEnd=19, url=null, language=null, rfNumber=[23], rfOrder=25, authorNames=WOO S, PARK J, LEE J Y, KWEON I S, journalName=null, refType=null, unstructuredReference=WOO S, PARK J, LEE J Y, KWEON I S, Cbam: convolutional block attention module[C]//Proceedings of the European Conference on Computer Vision (ECCV), 2018: 3-19., articleTitle=Cbam: convolutional block attention module, refAbstract=null), Reference(id=1312006337136001508, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, doi=null, pmid=null, pmcid=null, year=2019, volume=363, issue=null, pageStart=1, pageEnd=8, url=null, language=null, rfNumber=[24], rfOrder=26, authorNames=GAO J, WANG Q, journalName=Neurocomputing, refType=null, unstructuredReference=GAO J, WANG Q. SCAR: spatial-/channel-wise attention regression networks for crowd counting[J]. Neurocomputing, 2019, 363: 1-8., articleTitle=SCAR: spatial-/channel-wise attention regression networks for crowd counting, refAbstract=null)], funds=[Fund(id=1312006335047238089, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, awardId=112100000066, language=CN, fundingSource=喜洋洋公司农业科研发展基金项目(112100000066), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)]), AuthorCompany(id=1312006329678528889, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=2., ext=[AuthorCompanyExt(id=1312006329686917498, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China), AuthorCompanyExt(id=1312006329762414971, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.安徽喜洋洋农业科技有限公司,合肥 231521)])], figs=[ArticleFig(id=1312006332610347435, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 1, caption=Geographical location of study area and orthophoto, figureFileSmall=+FQspT3sq9ZM/VvRCVhFrg==, figureFileBig=sXgpXe+lroTyr1hvKIG4Rg==, tableContent=null), ArticleFig(id=1312006332706816428, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图1, caption=研究区域地理位置及正射影像, figureFileSmall=+FQspT3sq9ZM/VvRCVhFrg==, figureFileBig=sXgpXe+lroTyr1hvKIG4Rg==, tableContent=null), ArticleFig(id=1312006332887171501, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 2, caption=Structure diagram of rice counting network, figureFileSmall=Sd4wPGsmoG7rUARaurnifQ==, figureFileBig=IFQTo6fh1Js/27X67GOzWg==, tableContent=null), ArticleFig(id=1312006332975251886, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图2, caption=水稻苗计数网络结构图, figureFileSmall=Sd4wPGsmoG7rUARaurnifQ==, figureFileBig=IFQTo6fh1Js/27X67GOzWg==, tableContent=null), ArticleFig(id=1312006333038166447, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 3, caption=Specific structure of MCNN backbone for rice counting network, figureFileSmall=sZKuLr1m1XFiWYcE5QYuSg==, figureFileBig=i1a9sOGQS3RyEOCjSm9fng==, tableContent=null), ArticleFig(id=1312006333101081008, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图3, caption=水稻苗计数MCNN主干结构图, figureFileSmall=sZKuLr1m1XFiWYcE5QYuSg==, figureFileBig=i1a9sOGQS3RyEOCjSm9fng==, tableContent=null), ArticleFig(id=1312006333184967089, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 4, caption=Similarity-enhanced Gaussian kernel, figureFileSmall=czg4Bc2aNCRf57lN03dXSQ==, figureFileBig=5fLzGzt5cdtvszTM7iAWWg==, tableContent=null), ArticleFig(id=1312006333281436082, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图4, caption=相似度改进的高斯核处理效果, figureFileSmall=czg4Bc2aNCRf57lN03dXSQ==, figureFileBig=5fLzGzt5cdtvszTM7iAWWg==, tableContent=null), ArticleFig(id=1312006333361127859, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 5, caption=3×3 dilated convolution fusion Conv, figureFileSmall=83DMfHU/6Jk8ntJD2KBdtA==, figureFileBig=9Usz/lHWaMTyM1GEDwmsbQ==, tableContent=null), ArticleFig(id=1312006333457596852, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图5, caption=3×3膨胀卷积融合模块, figureFileSmall=83DMfHU/6Jk8ntJD2KBdtA==, figureFileBig=9Usz/lHWaMTyM1GEDwmsbQ==, tableContent=null), ArticleFig(id=1312006333520511413, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 6, caption=Model performance of ground truth generation using different Gaussian kernel, figureFileSmall=b/xo1qJcLPV3BTdTdIA4bQ==, figureFileBig=jUhle6pkdjh9EEIYCTQx9A==, tableContent=null), ArticleFig(id=1312006333591814582, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图6, caption=不同高斯核生成地面真值的模型性能, figureFileSmall=b/xo1qJcLPV3BTdTdIA4bQ==, figureFileBig=jUhle6pkdjh9EEIYCTQx9A==, tableContent=null), ArticleFig(id=1312006333650534839, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 7, caption=Model performance of using DCFC in different branches and layers, figureFileSmall=8AJcSOtTyH5bRnjXLh6Daw==, figureFileBig=FYNHWGUN8aaJy3f28DPHKQ==, tableContent=null), ArticleFig(id=1312006333742809528, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图7, caption=各个分支不同层使用DCFC的模型性能, figureFileSmall=8AJcSOtTyH5bRnjXLh6Daw==, figureFileBig=FYNHWGUN8aaJy3f28DPHKQ==, tableContent=null), ArticleFig(id=1312006333818307001, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 8, caption=Examples of different scenarios that affected counting accuracy, figureFileSmall=z/EfQsa9lzJblJzHD6xeEQ==, figureFileBig=XSRYzadBMl1wRVHR7nJIOw==, tableContent=null), ArticleFig(id=1312006333902193082, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图8, caption=影响计数精度的不同场景示例, figureFileSmall=z/EfQsa9lzJblJzHD6xeEQ==, figureFileBig=XSRYzadBMl1wRVHR7nJIOw==, tableContent=null), ArticleFig(id=1312006333977690555, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 9, caption=Mainstream counting networks (MCNN, CSRNet) and RCN density maps and counting results, figureFileSmall=iVs/vVfTu0mBkiW5nyd8FQ==, figureFileBig=d1mqYI0tbypSB7h55oiPxA==, tableContent=null), ArticleFig(id=1312006334044799420, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图9, caption=主流计数网络(MCNN、CSRNet)和RCN密度图以及计数结果, figureFileSmall=iVs/vVfTu0mBkiW5nyd8FQ==, figureFileBig=d1mqYI0tbypSB7h55oiPxA==, tableContent=null), ArticleFig(id=1312006334103519677, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Fig. 10, caption=Density distribution of rice field preparation, figureFileSmall=qtDzpJcHz1z4s4ne38YdfA==, figureFileBig=ZSL0PIGngInCGc2ijWwg5Q==, tableContent=null), ArticleFig(id=1312006334166434238, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=图10, caption=水稻整田的密度分布, figureFileSmall=qtDzpJcHz1z4s4ne38YdfA==, figureFileBig=ZSL0PIGngInCGc2ijWwg5Q==, tableContent=null), ArticleFig(id=1312006334225154495, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Tab. 1, caption=

Dataset

, figureFileSmall=null, figureFileBig=null, tableContent=
数据集图像数量水稻苗数/株平均每幅图像水稻苗数/株
训练集1 000173 241173.24
测试集30050 517168.39
合计1 300223 758172.12
), ArticleFig(id=1312006334288069056, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=表1, caption=

数据集参数

, figureFileSmall=null, figureFileBig=null, tableContent=
数据集图像数量水稻苗数/株平均每幅图像水稻苗数/株
训练集1 000173 241173.24
测试集30050 517168.39
合计1 300223 758172.12
), ArticleFig(id=1312006334355177921, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Tab. 2, caption=

Receptive field size of different branches and layers after dilated convolution

, figureFileSmall=null, figureFileBig=null, tableContent=
分支卷积/池化核(k)步数感受野大小
Branch 1Conv(9)19
Pooling(2)110
DCFC(7/13)116/22
Pooling(2)117/23
DCFC(7/13)123/29/30/35
Conv(7)129/35/36/41
Branch 2Conv(7)19
Pooling(2)110
Conv(5)114
Pooling(2)115
Conv(5)119
Branch 3Conv(5)15
Pooling(2)16
Conv(5)110
Pooling(2)111
Conv(5)115
), ArticleFig(id=1312006334426481090, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=表2, caption=

膨胀卷积融合后不同分支和层的感受野大小

, figureFileSmall=null, figureFileBig=null, tableContent=
分支卷积/池化核(k)步数感受野大小
Branch 1Conv(9)19
Pooling(2)110
DCFC(7/13)116/22
Pooling(2)117/23
DCFC(7/13)123/29/30/35
Conv(7)129/35/36/41
Branch 2Conv(7)19
Pooling(2)110
Conv(5)114
Pooling(2)115
Conv(5)119
Branch 3Conv(5)15
Pooling(2)16
Conv(5)110
Pooling(2)111
Conv(5)115
), ArticleFig(id=1312006334497784259, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Tab. 3, caption=

Performance under different attention model

, figureFileSmall=null, figureFileBig=null, tableContent=
模块MAE/(穴·m-2)MSE/(穴2·m-4)rMAE/%
None5.746.153.80
SAM4.206.713.60
CAM5.126.253.52
CBAM4.936.103.20
SAM+CAM+CBAM3.104.032.90
), ArticleFig(id=1312006334564893124, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=表3, caption=

不同注意力模块下模型性能

, figureFileSmall=null, figureFileBig=null, tableContent=
模块MAE/(穴·m-2)MSE/(穴2·m-4)rMAE/%
None5.746.153.80
SAM4.206.713.60
CAM5.126.253.52
CBAM4.936.103.20
SAM+CAM+CBAM3.104.032.90
), ArticleFig(id=1312006334686527941, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Tab. 4, caption=

Model results based on different attention modules in specific scenarios

, figureFileSmall=null, figureFileBig=null, tableContent=
场景模块MAE/(穴·m-2)MSE/(穴2·m-4)rMAE/%
复杂背景CAM5.157.163.45
SAM10.1210.125.11
None10.0113.5812.15
多尺度CAM7.158.215.10
SAM4.915.153.66
None13.219.355.68
密集遮挡CAM6.158.147.15
SAM8.199.936.15
None11.1913.9310.15
), ArticleFig(id=1312006334762025414, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=表4, caption=

不同注意力模块在特定场景下的模型结果

, figureFileSmall=null, figureFileBig=null, tableContent=
场景模块MAE/(穴·m-2)MSE/(穴2·m-4)rMAE/%
复杂背景CAM5.157.163.45
SAM10.1210.125.11
None10.0113.5812.15
多尺度CAM7.158.215.10
SAM4.915.153.66
None13.219.355.68
密集遮挡CAM6.158.147.15
SAM8.199.936.15
None11.1913.9310.15
), ArticleFig(id=1312006334862688711, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=EN, label=Tab. 5, caption=

Results of several classic density map counting models and RCN models

, figureFileSmall=null, figureFileBig=null, tableContent=
模型MAE/(穴·m-2)MSE/(穴2·m-4)rMAE/%
MCNN11.1713.8312.54
CSRNet10.2511.2410.22
SCAR13.2915.2910.27
TasselNetv27.318.236.21
RCN(本文)3.104.032.90
), ArticleFig(id=1312006334925603272, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, language=CN, label=表5, caption=

几种经典密度图计数模型和RCN的模型结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型MAE/(穴·m-2)MSE/(穴2·m-4)rMAE/%
MCNN11.1713.8312.54
CSRNet10.2511.2410.22
SCAR13.2915.2910.27
TasselNetv27.318.236.21
RCN(本文)3.104.032.90
)], attaches=null, journal=Journal(id=1311384913490051186, delFlag=0, nameCn=农业机械学报, nameEn=Transactions of the Chinese Society for Agricultural Machinery, nameHistory1=null, nameHistory2=null, issn=1000-1298, eissn=null, cn=11-1964/S, coden=CODEN NUYCA3, 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=DpMljKbCTPO4JEohCaKm5A==, journalPrice=null, startedYear=null, abbrevIsoEn=Transactions of the Chinese Society for Agricultural Machinery, journalRemark=null, publicationField=null, createdTime=1790588127731, updatedTime=1790740716593, createdBy=18614031015, updatedBy=18614031015, firstLetterCn=N, firstLetterEn=N, subjectCode=Agriculture, subjectName=null, subjectCodeEn=Agriculture, subjectNameEn=null, picCn=DpMljKbCTPO4JEohCaKm5A==, picEn=klDuz9nOYH3sS2mMFZcDKQ==, jcr=null, cjcr=null, exts=[JournalExt(id=1311386928897945885, language=CN, name=农业机械学报, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1790588608240, updatedTime=1790588608240, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://www.j-csam.org/jcsam/author/login, submissionEditorUrl=https://www.j-csam.org/jcsam/editor/login, submissionReviewUrl=https://www.j-csam.org/jcsam/user/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1311386928944083230, language=EN, name=Transactions of the Chinese Society of Agricultural Machinery, 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=1790588608251, updatedTime=1790588608251, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://www.j-csam.org/jcsam/author/login, submissionEditorUrl=https://www.j-csam.org/jcsam/editor/login, submissionReviewUrl=https://www.j-csam.org/jcsam/user/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1311385739772133495, websiteList=[Website(id=1311386135197221186, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1311385739772133495, 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/nyjxxb/CN, language=CN, createTime=1790588419007, createBy=18614031015, updateTime=1790588455647, updateBy=18614031015, name=农业机械学报-中文, tplId=1146099689490845704, title=农业机械学报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1312024622120399707, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=articleTextType, value=kx, createTime=1790740646152, updateTime=1790740646152, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622091039576, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=banner, value=null, createTime=1790740646145, updateTime=1790740646145, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622149759838, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=grayFlag, value=0, createTime=1790740646159, updateTime=1790740646159, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622078456663, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=logo, value=https://castjournals.cast.org.cn/joweb/nyjxxb/CN/file/pic?fileId=oEf7LShi6rv3EG1pNx7PwQ==, createTime=1790740646142, updateTime=1790740646142, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622166537056, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=minRunFlag, value=0, createTime=1790740646163, updateTime=1790740646163, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622112011098, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/nyjxxb/CN/file/pic, createTime=1790740646150, updateTime=1790740646150, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622158148447, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=silenceFlag, value=0, createTime=1790740646161, updateTime=1790740646161, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622099428185, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1790740646147, updateTime=1790740646147, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622128788316, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=themeColor, value=null, createTime=1790740646154, updateTime=1790740646154, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024622141371229, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135197221186, code=themeStyle, value=null, createTime=1790740646157, updateTime=1790740646157, creator=18614031015, updator=18614031015)]), Website(id=1311386135398547788, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1311385739772133495, 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/nyjxxb/EN, language=EN, createTime=1790588419055, createBy=18614031015, updateTime=1790740692575, updateBy=18614031015, name=农业机械学报-英文, tplId=1146101810881728533, title=Transactions of the Chinese Society for Agricultural Machinery, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1312024677325828967, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=articleTextType, value=kx, createTime=1790740659314, updateTime=1790740659314, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677300663140, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=banner, value=null, createTime=1790740659308, updateTime=1790740659308, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677350994794, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=grayFlag, value=0, createTime=1790740659320, updateTime=1790740659320, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677288080227, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=logo, value=https://castjournals.cast.org.cn/joweb/nyjxxb/EN/file/pic?fileId=oEf7LShi6rv3EG1pNx7PwQ==, createTime=1790740659305, updateTime=1790740659305, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677367772012, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=minRunFlag, value=0, createTime=1790740659324, updateTime=1790740659324, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677317440358, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/nyjxxb/EN/file/pic, createTime=1790740659312, updateTime=1790740659312, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677359383403, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=silenceFlag, value=0, createTime=1790740659322, updateTime=1790740659322, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677309051749, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1790740659310, updateTime=1790740659310, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677334217576, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=themeColor, value=null, createTime=1790740659316, updateTime=1790740659316, creator=18614031015, updator=18614031015), WebsiteProps(id=1312024677342606185, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1311386135398547788, code=themeStyle, value=null, createTime=1790740659318, updateTime=1790740659318, creator=18614031015, updator=18614031015)])], journalTitle=农业机械学报, weixinUrl=null, journalUrl=https://www.j-csam.org/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Transactions of the Chinese Society for Agricultural Machinery, journalPhotoCn=DpMljKbCTPO4JEohCaKm5A==, journalPhotoEn=klDuz9nOYH3sS2mMFZcDKQ==, journalFirstLetter=N, 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/nyjxxb/CN/10.6041/j.issn.1000-1298.2025.12.051, detailUrlEn=https://castjournals.cast.org.cn/joweb/nyjxxb/EN/10.6041/j.issn.1000-1298.2025.12.051, pdfUrlCn=https://castjournals.cast.org.cn/joweb/nyjxxb/CN/PDF/10.6041/j.issn.1000-1298.2025.12.051, pdfUrlEn=https://castjournals.cast.org.cn/joweb/nyjxxb/EN/PDF/10.6041/j.issn.1000-1298.2025.12.051, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1766592000000, fullTextJson=null, articleText=null, reference=null)
收藏切换
融合多尺度特征与多列卷积网络的水稻苗无人机遥感计数方法
收藏切换
PDF下载
潘义 1 , 苏浩然 1 , 秦琳琳 1 , 吴刚 1 , 石春 1 , 王宜坤 2 , 华胜 2
农业机械学报 | 农业信息化工程 2025,56(12): 560-567
收起
收藏切换
农业机械学报 |农业信息化工程 2025 , 56 (12) : 560 -567
融合多尺度特征与多列卷积网络的水稻苗无人机遥感计数方法
全屏
[Author(id=1312006329837912445, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=alexpanyi@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1312006329909215615, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006329837912445, language=EN, stringName=Yi PAN, firstName=Yi, middleName=null, lastName=PAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006329972130176, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006329837912445, language=CN, stringName=潘义, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio={"content":"

潘义(2000—),男,硕士生,主要从事农业信息化技术研究,E-mail:

"}, bioImg=null, bioContent=

潘义(2000—),男,硕士生,主要从事农业信息化技术研究,E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006330047627650, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, 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=1312006330127319428, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330047627650, language=EN, stringName=Haoran SU, firstName=Haoran, middleName=null, lastName=SU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006330215399813, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330047627650, language=CN, stringName=苏浩然, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006330303480199, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=qinll@ustc.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1312006330404143497, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330303480199, language=EN, stringName=Linlin QIN, firstName=Linlin, middleName=null, lastName=QIN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006330475446666, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330303480199, language=CN, stringName=秦琳琳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006330571915660, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, 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=1312006330664190350, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330571915660, language=EN, stringName=Gang WU, firstName=Gang, middleName=null, lastName=WU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006330731299215, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330571915660, language=CN, stringName=吴刚, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006330806796689, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, 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=1312006330899071379, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330806796689, language=EN, stringName=Chun SHI, firstName=Chun, middleName=null, lastName=SHI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006330987151764, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006330806796689, language=CN, stringName=石春, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国科学技术大学自动化系,合肥 230026, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329586254198, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=1., ext=[AuthorCompanyExt(id=1312006329594642807, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.Department of Automation, University of Science and Technology of China, Hefei 230026, China), AuthorCompanyExt(id=1312006329607225720, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329586254198, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国科学技术大学自动化系,合肥 230026)])]), Author(id=1312006331071037846, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, 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=1312006331150729624, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006331071037846, language=EN, stringName=Yikun WANG, firstName=Yikun, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006331213644185, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006331071037846, language=CN, stringName=王宜坤, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.安徽喜洋洋农业科技有限公司,合肥 231521, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329678528889, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=2., ext=[AuthorCompanyExt(id=1312006329686917498, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China), AuthorCompanyExt(id=1312006329762414971, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.安徽喜洋洋农业科技有限公司,合肥 231521)])]), Author(id=1312006331280753051, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, orderNo=6, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1312006331352056221, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006331280753051, language=EN, stringName=Sheng HUA, firstName=Sheng, middleName=null, lastName=HUA, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1312006331456913822, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, authorId=1312006331280753051, language=CN, stringName=华胜, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2.安徽喜洋洋农业科技有限公司,合肥 231521, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1312006329678528889, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, xref=2., ext=[AuthorCompanyExt(id=1312006329686917498, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China), AuthorCompanyExt(id=1312006329762414971, tenantId=1146029695717560320, journalId=1311385739772133495, articleId=1312006326264365397, companyId=1312006329678528889, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.安徽喜洋洋农业科技有限公司,合肥 231521)])])]
潘义1 , 苏浩然1, 秦琳琳1 , 吴刚1, 石春1, 王宜坤2, 华胜2
作者信息
  • 1.中国科学技术大学自动化系,合肥 230026
  • 2.安徽喜洋洋农业科技有限公司,合肥 231521
通讯作者:
秦琳琳(1975—),女,高级工程师,博士,主要从事人工环境建模与控制研究,E-mail:
作者简介:

潘义(2000—),男,硕士生,主要从事农业信息化技术研究,E-mail:

Rice Seedling Counting Method Based on Multi-column Convolutional Network Integrating Multi-scale Features
Yi PAN1 , Haoran SU1, Linlin QIN1 , Gang WU1, Chun SHI1, Yikun WANG2, Sheng HUA2
Affiliations
  • 1.Department of Automation, University of Science and Technology of China, Hefei 230026, China
  • 2.Anhui Xiyangyang Agricultural Technology Co., Ltd., Hefei 231521, China
出版时间: 2025-12-25 doi: 10.6041/j.issn.1000-1298.2025.12.051
文章导航
收藏切换

水稻早苗期种植密度与产量估算、田间管理密切相关。随着计算机视觉技术的发展,利用无人机图像识别技术自动统计水稻数量正逐步取代传统的人工采样和计数方法。然而,无人机拍摄的图像存在水稻苗形状不规则、水稻苗重叠、尺度变化和背景遮挡等问题,导致特征学习出现偏差,进而影响了水稻密度的准确估算和计算预测的效率。水稻苗之间显著的相对大小差异也成为影响计数精度的另一重要因素。针对这些问题,提出了一种基于改进的多列卷积神经网络的水稻苗计数算法(Rice counting network,RCN)。该算法根据目标的大小和形状,提出了改进高斯密度核,结合水稻图像与中心点RGB图像的相似度信息,使密度核更加贴合水稻苗纹理。对不同大小和形状的稻穗数量进行量化,并对尺度进行精细划分,计算出需要融合的不同特征感受野大小,采用融合膨胀卷积模块,增加了感受野大小,增强了模型对多尺度环境中的处理效果。在此基础上,对不同尺度的通路添加合适的注意力机制,增强不同尺度下对水稻苗合理信息的关注,在复杂背景、多尺度和密集遮挡的场景中体现出了明显的有效性。在高度5 m捕获的RSD2023数据集上,RCN的平均绝对误差为3.10穴/m2,均方误差为4.03穴2/m4,优于MCNN、CSRNet、SCAR和TasselNetV2等主流算法。

水稻苗  /  计数  /  密度图  /  多列神经网络  /  无人机遥感  /  多尺度特征

The planting density of rice in the early seedling stage is closely related to yield estimation and field management. With the development of computer vision technology, the use of drone image recognition technology to automatically count rice is gradually replacing traditional manual sampling and counting methods. However, images taken by drones have problems such as irregular rice seedling shapes, overlapping rice seedlings, scale changes, and background occlusion, which lead to deviations in feature learning, which in turn affects the accurate estimation of rice density and the efficiency of computational prediction. Significant relative size differences among rice seedlings will also become another important factor affecting counting accuracy. In response to these problems, a rice seedling counting algorithm (rice counting network, RCN) was proposed based on an improved multi-column convolutional neural network. Based on the size and shape of the target, an improved Gaussian density kernel was proposed, which combined the similarity information of the rice image and the RGB image of the center point to make the density kernel more suitable for the texture of the rice seedlings. The number of rice ears of different sizes and shapes was quantified, and the scales were divided finely to calculate the receptive field sizes of different features that needed to be fused. The fusion expansion convolution module was used to increase the receptive field size and enhance the model's multi-scale capabilities. On this basis, appropriate attention mechanisms were added to the pathways at different scales to enhance attention to reasonable information about rice seedlings at different scales. It showed obvious effectiveness in scenes with complex backgrounds, multi-scales and dense occlusions. On the RSD2023 data set captured at a height of 5 m, RCN's mean absolute error MAE=3.10 holes/m2 and mean square error MSE=4.03 holes2/m4, which were better than that of the mainstream algorithms such as MCNN, CSRNet, SCAR and TasselNetV2.

rice seedlings  /  counting  /  density map  /  multi-column neural network  /  UAV remote sensing  /  multi-scale feature
潘义, 苏浩然, 秦琳琳, 吴刚, 石春, 王宜坤, 华胜. 融合多尺度特征与多列卷积网络的水稻苗无人机遥感计数方法. 农业机械学报, 2025 , 56 (12) : 560 -567 . DOI: 10.6041/j.issn.1000-1298.2025.12.051
Yi PAN, Haoran SU, Linlin QIN, Gang WU, Chun SHI, Yikun WANG, Sheng HUA. Rice Seedling Counting Method Based on Multi-column Convolutional Network Integrating Multi-scale Features[J]. Transactions of the Chinese Society for Agricultural Machinery, 2025 , 56 (12) : 560 -567 . DOI: 10.6041/j.issn.1000-1298.2025.12.051
我国是最大的水稻生产国,水稻产量占我国粮食总产量的30%[1]。在水稻种植过程中,种植密度是影响水稻生长和产量的重要因素[2]。水稻苗期的植株密度对预测产量、最终成穗密度及生长状况具有决定性作用,显著影响作物生长的多个方面及田间管理措施,是实现籽粒高产的重要农艺因素[3]。
因此,水稻苗期种植密度的计算至关重要。目前,水稻苗期植株计数仍然主要依赖于实地调查的人工采样和计数,并且经常需要人工处理每个样本,过程繁琐、低效且不准确[4]。
随着无人机遥感技术的不断进步,无人机平台在农业中的应用越来越广泛。无人机作为一种低成本的替代检测和数据分析技术[5],在农业信息采集中具有灵活、经济、周期短等特点[6]。在水稻监测方面,无人机具有快速获取植被信息和实时动态监测的优势。
深度学习在图像分类、目标检测和图像分割方面展现出了优异的性能[7],并已应用于植物计数[8-9]。然而,标注图像是一项非常耗费人力和时间的任务,这通常导致全面注释的数据集包含较少的图像,而具有更多图像的数据集则注释不够详细[10]。此外,遮挡和重叠会导致模型的检测和计数结果出现严重偏差[11]。
密度图回归是一种用于估计图像中物体密度或数量的技术。它通过将检测和回归方法相结合,在提高召回率的同时输出物体的位置信息,该方法在物体计数领域有着广泛的应用[12-18]。
然而,基于多列卷积神经网络(Multi-column convolutional network,MCNN)的植物计数研究相对较少,大多数研究是基于简单背景下的单一尺度物体计数[19-20]。
当水稻苗的分辨率和绝对尺寸增加时,MCNN的最大感受野不足以有效提取目标特征,导致感受野与多尺度物体大小之间不匹配,TasselNetV2[16]模型通过使用上下文信息来挖掘对象与像素之间的关系,从而扩展感受野,但这会显著增加网络的计算开销。因此,在MCNN模型的基础上,如何调整感受野以适应实际的多尺度水稻苗,是将MCNN应用于水稻苗计数需解决的关键问题。
农业领域中的计数对象是植物,其形状往往不是规则的圆形,导致由高斯核生成的密度图与实际植被分布不完全一致。文献[21]采用基于协方差矩阵改进的高斯核来模拟稻穗等狭长物体的位置分布,但这种方法仍存在局限性,即高斯核的固定性使其无法根据植被具体形状变化灵活调整。
水稻田还具有复杂背景及密集遮挡等问题。文献[22]发现,视觉注意机制在识别相似和复杂背景中的物体方面表现良好,为解决农田场景中水田反光、水稻苗遮挡、尺度差异等视觉问题提供了思路。受人眼主动动态感知信息模型的启发,结合区域注意力机制的应用,可以提高对密集场景、复杂背景以及尺度多样化场景下水稻苗检测的精度。
综上,针对高密度水稻苗计数的问题,本文通过改进多列卷积神经网络,采用相似度改进密度图,并使用膨胀卷积改进卷积核以增加感受野,采用合适的注意力机制对不同尺度分支进行细节和整体信息的增强,建立水稻苗计数网络(Rice counting network,RCN)。
实验在安徽省合肥市庐江县实验基地(31°47′N,117°23′E)开展,如图1所示。实验区属于亚热带季风气候区,四季分明,气温变化明显,主要作物为小麦和水稻。全县年平均气温为16℃,年平均降水量为1 170.5 mm。降水主要集中在4—8月,7月达到峰值。本次实验所用的水稻品种为常规晚粳稻品种银香38。
实验中使用大疆M100型无人机,在高度5 m处以2 m/s的速度采集数据,水平重叠率为80%,纵向重叠率为80%。无人机图像于2023年8月8日前后水稻的幼苗期拍摄。数据收集时间主要在当地10:00—13:00,在有利的光照条件下进行。使用大疆GS PRO软件进行航路规划,确保每次数据采集的高度和区域保持一致。
大疆M100型无人机搭载ZENMUSE X3型相机,分辨率为1 920像素×1 080像素,在5 m空中的分辨率约为0.3 cm,采用1 s间隔拍摄,并保存为jpg格式。最终剔除相似度过高或清晰度较差的图像,并使用Photoscan软件进行对齐,生成相应的正射影像。
由于水稻苗是具有中心特征的植物,本文采用点标注的方法,对水稻苗的中心进行标注,因为点标注只需要对其中心进行标注,故标注速度快,节省人力成本。使用Python编写程序来标注水稻植株的中心,并以左上角坐标为原点记录每株水稻的中心坐标,然后生成相应的.mat文件。数据集的详细信息如表1所示,命名为RSD2023(Rice Seeding Dataset 2023)。
为了防止光照强弱、拍摄角度等对训练产生影响,对图像进行旋转、裁剪、随机改变亮度和顺序打乱处理,增加了数据的鲁棒性。
本文以多列卷积神经网络(MCNN)为主干框架建立RCN。MCNN在高密度、高遮挡的人群计数场景中得到了较好的密度图和计数精度。本研究基于水稻株种植的场景对MCNN网络做了以下改进:
(1)在密度图生成过程中,引入了相似度增强的高斯核,通过将水稻苗周围像素点与中心点的相似度,与高斯卷积核进行点对点相乘,解决了由于水稻苗的不规则形状而导致的高斯核圆形不匹配问题。
(2)增加了膨胀卷积融合模块,与原始特征层融合,构建了多尺度的特征层,以解决多尺度稻田环境下MCNN感受野范围过小的问题。
(3)针对MCNN的3个感受野分支,分别引入了不同的注意力机制,增强对复杂稻田环境下复杂背景和遮挡问题的信息关注度。
RCN结构如图2所示,由3个模块组成:密度图生成模块(Density map generation module,DMFM)、改进的MCNN主干(Improved multi-column convolutional backbone,IMCB)、后处理模块(Post-processing module,PM)。
其中改进的MCNN主干结构如图3所示,共包含膨胀卷积、池化层、普通卷积和注意力模块4个模块。为了估计图像中的水稻苗数量,首先IMCB中通过不同尺度的卷积获取不同尺度特征,接着采用膨胀卷积扩展网络的感受野,然后使用注意力机制关注不同尺度中更加关键的信息,最后通过1×1卷积将这些特征融合,生成相应的密度图,通过PM模块进行分辨率的恢复并回归完成水稻苗计数。
在模型中的DMFM模块,首先按照类似MCNN的方法,根据标注的水稻株中心坐标进行高斯核卷积,生成从中心点向周围逐渐减小的概率密度图,即地面真值(Ground truth,GD)。高斯核卷积核公式为
其中
式中
      H(x,y)——对应(x,y)坐标高斯核的值
      (x,y)——标注点坐标
      (xi,yi)——标注点周围点坐标
      σ——标准差
从图4中可以看出,水稻株呈现出不规则形状,而非标准的圆形,与MCNN模型中的人群计数特征不同。针对这一问题,在高斯核生成地面真值后,计算原始图像中心点与周围点的相似度,并将其与高斯卷积核进行点对点相乘生成地面真值,以增强密度图对细节信息的提取,确保地面真值沿水稻苗纹理从中心向外逐渐减小。生成相似度增强的高斯核(Similarity-enhanced Gaussian kernel,SGK)的公式为
式中
      G(x,y)——对应(x,y)的真实密度图
      m(x,y)——对应图像像素灰度
      (a,b)——标注的水稻株的中心点坐标
通过上述处理,在不规则的水稻株中,能更好地提取密度图的细节信息,其效果如图4所示。
在模型中的IMCB模块中,为了解决多尺度稻田环境下MCNN感受野范围过小的问题,引入了一种基于膨胀卷积融合的模块(Dilated convolution fusion conv,DCFC)。以3×3卷积模型为例,具体方法如图5所示。
分别应用3×3的标准卷积和膨胀率为2的3×3膨胀卷积对输入进行卷积处理,然后将它们的输出进行通道叠加,从图5中可以看出3×3的DCFC公式为对应卷积的感受野大小分别为3×3和5×5,对于其他大小为n×n的卷积核,膨胀率为2后的感受野大小分别为n×n和(2n-1)×(2n-1),因此,膨胀卷积后的感受野从n×n扩展到了(2n-1)×(2n-1)。
为了保持输出图像尺寸不变,需要对图像周围进行填充,填充大小的计算公式为
式中
      p——填充大小
      d——膨胀率
      i——输入图像尺寸
在模型中的IMCB模块中,传统的MCNN模型通过3种不同尺度的分支来获取不同尺度的目标特征,但是对于不同尺度信息的重视程度是相同的。然而,在实际的水稻株特征提取过程中,不同尺度的信息关注点是不同的。为了提高模型在计数精度上的表现,本文引入了3种不同的注意力机制,其中融合空间注意力和通道注意力机制的卷积块注意力模块(Convolutional block attention module,CBAM)[23]适用于中尺度范围。为了减少模型的资源消耗和运行时间,将空间注意力和通道注意力分别应用于更加适用的尺度上。在小尺度分支中,引入了通道注意力机制(Channel attention module,CAM),CAM能够有效应对高度拥挤的场景,其中前景纹理与一些背景区域(如树木、建筑等)非常相似,CAM的应用能够有效减小估计误差[24]。同时,在大尺度分支中,加入了空间注意力机制(Spatial attention module,SAM),SAM能够在空间维度上捕捉广泛的依赖关系,特别适用于发现具有相似局部图案和纹理特征的场景,因此将其应用于尺度变化范围较大的场景中。
本文使用平均绝对误差(Mean absolute error,MAE)、均方误差(Mean square error,MSE)、相对平均绝对误差(Root mean absolute error,rMAE)来评价本模型的计数性能。
本实验在64位Windows 11操作系统的台式机上进行,运行内存为16 GB,采用RTX 3060 GPU进行训练,模型使用PyTorch框架,Python版本为3.11,完成了水稻苗技术网络的构建。
训练参数设置如下:周期数为2 000,学习率为0.000 01,批处理大小为500,动量(Momentum)设置为0.9,优化算法为亚当(Adam)优化算法。
SGK生成的主要参数有:高斯核标准差σ、高斯核大小s×s、相似度的权重α。对于高度5 m拍摄的水稻苗图像进行统计,其在截取的1 000像素×1 000像素分辨率图像中,水稻苗的平均尺度大概在20像素×20像素到40像素×40像素之间,平均大小约为900像素。
为了验证本文采用相似性和高斯核结合的密度图生成方法的有效性,使用原始MCNN模型。对比原始最佳参数与SGK在水稻苗密度图生成中的精度。最终结果如图6所示,可以得出以下结论:本文采用原始标准卷积核训练MCNN,通过对比实验选择最符合图像中水稻苗大小的σ和s,其中变量σ∈{5,10,15,20},s×s∈{30×30,40×40,50×50}。对比实验可以发现,当高斯核和水稻苗大小比较接近时,在不同的σ取值中预测效果都较好。通过对比实验,取σ=15,s=40时,MAE最小降至11.17穴/m2,MSE最小降至13.83穴2/m4。因此选择σ=15,s=40作为卷积核的参数。
在对比取得的高斯核卷积参数的基础上,采用SGK进一步改进高斯核。由于水稻苗的非标准圆形形状,通过SGK改进后,改进高斯核生成的密度图的真值可以沿着水稻苗的纹理向外扩散。从图6可以看出,采用SGK后,不同尺寸的高斯核在MCNN中的表现都有所提升,当σ=15,s=40时,MSE取得最小值9.17穴2/m4。在后续实验中,始终采用σ=15,s=40的SGK生成地面真值。
本文在MCNN模型的基础上添加了DCFC模块,增加了其感受野而不增加额外计算量。为了确定DCFC在网络结构中的合适位置,采用对比方法选择最佳安放位置。MCNN模型共有3个分支,大尺度分支Branch 1、中尺度分支Branch 2和小尺度分支Branch 3,进入每个分支之后,均经过3个卷积核进行特征提取。通过将不同位置的卷积核替换成DCFC来进行对比实验,每个分支中对应的数字表示进入该分支后第几个卷积核替换成DCFC,实验结果如图7所示。在Branch 1中使用DCFC时,与原始模型相比,MAE都减少了,且MSE的降低幅度相对MAE和rMAE的减少幅度更大,当2、3层使用DCFC时,模型的精度最高。
在Branch 2和Branch 3中使用DCFC时,MAE、MSE和rMAE都没有显著的下降,甚至出现上升的情况,说明在中尺度和小尺度中使用DCFC的效果不好,对水稻苗计数的预测效果不好。
当选取上述实验中效果最好的组合,即Branch 1中的第2、3层卷积替换成DCFC,计算了各个分支中每一层的感受野大小,如表2所示。根据表2中的感受野可以看出,感受野最大达到41×41。从上文中得知水稻苗的尺度大概在20像素×20像素到40像素×40像素之间,可以看出DCFC模块加强了模型对大水稻苗的感受,但是当Branch 1中3层都使用DCFC的时候,通过计算,最大感受野达到了47×47,超出了本实验中最大的水稻苗尺寸,造成了信息的冗余。对Branch 2和Branch 3中的每一层添加DCFC时,可以看到相对于不添加DCFC模块,某些组合的MSE减小,但是基本上所有的rMAE都增加。从中可以得知,DCFC可以增加每个卷积的感受野,但对每个感受野的预测精度下降。
因此Branch 1中采用DCFC很好地扩大了感受野,感受到了尺度较大的水稻苗信息。说明当感受野增加时,模型对极端误差的处理效果更好。
对注意力模块进行对比实验,验证算法中注意力模块的有效性,实验结果如表3所示。
同时使用SAM、CAM和CBAM后,模型的MAE达到了3.10穴/m2,与各部分单独使用模块相比有显著降低,MSE下降得更加明显,可以看出注意力机制对较大误差的部分处理效果较好。
为了验证各个注意力机制在视觉问题上的应用效果,在RSD2023中选择对应场景(密集遮挡、复杂背景、多尺度)图像各50幅来对比并验证在不同场景下通道注意力机制的有效性,示例如图8所示,对数据集中对应场景进行分类,并对每种场景进行单独实验,来验证不同分支中注意力机制的有效性,实验结果如表4所示,在不同分支中加入不同注意力机制后,都不同程度上提高了计数精度,CAM在复杂背景和密集场景中的效果更加显著,SAM在多尺度场景下注意力模块的下降程度更高。说明了CAM和SAM解决复杂背景、密集遮挡和多尺度问题的有效性。
采用数据集RSD2023,本文选取当前主流的目标计数方法MCNN、CSRNet、SCAR、TasselNetV2进行对比研究,具体的参数设置均按照源码中的默认设置,结果如表5所示。
通过对网络结果的分析,可以看出RCN的综合误差更低,与TasselNetV2相比,本文模型RCN的MAE、MSE、rMAE分别降低了4.21穴/m2、4.20穴2/m4、3.31个百分点。
图9为随机选取的5幅图像,从图中可以看出,RCN对于水稻苗的中心识别更加明显,同时通过密度图计算出的水稻苗穴数更接近真实值。
为了保证本文的RCN算法在整块大田中的适用性,本文采用对不同区域的图像进行识别,并根据图像的GPS定位对图像生成的密度值进行拼接,颜色越深代表种植越密集,最终效果图如图10所示,很好地展示了水稻种植密度的分布情况。
对密度图进行后处理,计算出整田的水稻苗数量共为15 035穴。整田共0.71 hm2,机插时的行穴距为25 cm×16 cm,机插有效率为0.95,可以计算出理论水稻苗的穴数为15 841.25,与根据本实验计算出来的数量总体相近。
(1)通过SGK生成的地面真值可以提高模型对水稻苗细节纹理的提取,使输出的图像更加贴合水稻苗的细节,可以使模型对水稻苗的提取更加准确。本文SGK主要基于相似度设计了一种精细化的高斯核。对于无人机在高度5 m上成像的水稻苗场景,采用大小为40、σ为15的改进高斯核其MSE能够降至9.17穴2/m4,与采用原始MCNN中的高斯核训练模型相比,MAE降低了16.21穴/m2。
(2)DCFC可以增加模型的感受野,对不同尺度目标的鲁棒性更强,针对大田水稻苗期场景中,由于生长状况不一致,导致植物大小变化较多。为了增加对不同尺度的水稻苗的信息获取,本文根据MCNN原始架构,计算出了其对应各层各模块的感受野,通过DCFC模块使感受野的范围更符合水稻苗尺度范围,通过实验发现可以使得MSE从9.17穴2/m4降至5.74穴2/m4。
(3)针对背景复杂、密集遮挡、多尺度问题,在大尺度上使用通道注意力机制,提取更大尺度的水稻苗植株的上下文信息,并在小尺度上增加空间注意力机制,增加对较小水稻苗和复杂背景(水田反光、杂草、土壤)的识别精度。
(4)采用以上模块,田间水稻苗计数的精度得到了显著提高,其中精度最高MAE可以达到3.10穴/m2,MSE可以达到4.03穴2/m4,与MCNN、CSRNet、SCAR、TasselNetv2模型相比,MAE分别降低8.07、7.15、10.19、4.21穴/m2。MSE分别降低了9.80、7.21、11.26、4.20穴2/m4。自制数据集的实验结果表明,本文提出的RCN模型在处理高密度水稻苗计数问题上的有效性。
  • 喜洋洋公司农业科研发展基金项目(112100000066)
参考文献 引证文献
排序方式:
[1]
国家统计局. 中国统计年鉴2023[M]. 北京: 中国统计出版社, 2023.
[2]
梁青铎. 施氮量和机插密度对水稻生长、产量和氮肥利用效率的影响[D]. 北京: 中国农业科学院, 2020.
LIANG Qingduo. Effects of nitrogen application rate and planting density on growth, yield and nitrogen use efficiency of rice[D]. Beijing: Chinese Academy of Agricultural Sciences, 2020. (in Chinese)
[3]
ALIPOUR F, MOBASSER H R. Effect of planting density on growth characteristics and grain yield increase in successive cultivations of two rice cultivars[J]. Agrosystems, Geosciences & Environment, 2021, 4(4): 20213.
[4]
PICKSON R, GUI P, CHEN A, et al. Climate change and food security nexus in Asia: a regional comparison[J]. Ecological Informatics, 2023, 76: 102038.
[5]
ASLAN M, DURDU A, SABANCI K, et al. A comprehensive survey of the recent studies with UAV for precision agriculture in open fields and greenhouses[J]. Applied Sciences, 2022, 12(3): 1047.
[6]
ZHANG C, KOVACS J M. The application of small unmanned aerial systems for precision agriculture: a review[J]. Precision Agriculture, 2012, 13: 693-712.
[7]
KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.
[8]
KITANO B T, MENDES C C, GEUS A R, et al. Corn plant counting using deep learning and UAV images[J]. IEEE Geoscience and Remote Sensing Letters, 2019(8): 1-5.
[9]
王宏乐, 叶全洲, 王兴林, 等. 基于YOLOv7的无人机影像稻穗计数方法研究[J]. 广东农业科学, 2023, 50(7): 74-82.
WANG Hongle, YE Quanzhou, WANG Xinglin, et al. Rice panicles counting method based on YOLOv7 using unmanned aerial vehicles images[J]. Guangdong Agricultural Sciences, 2023, 50(7): 74-82. (in Chinese)
[10]
HANBURY A. A survey of methods for image annotation[J]. Journal of Visual Languages & Computing, 2008, 19(5): 617-627.
[11]
WANG Y, QIN Y, CUI J, et al. Occlusion robust wheat ear counting algorithm based on deep learning[J]. Frontiers in Plant Science, 2021, 12: 645899.
[12]
LEMPITSKY V, ZISSERMAN A. Learning to count objects in images[J]. Advances in Neural Information Processing Systems (NIPS), 2010, 1: 1324-1332.
[13]
ZHANG Y, ZHOU D, CHEN S, et al. Single-image crowd counting via multi-column convolutional neural network[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, 2016: 589-597.
[14]
LI Y, ZHANG X, CHEN D, et al. CSRNet: dilated convolutional neural networks for understanding the highly congested scenes[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, 2018: 1091-1100.
[15]
ZHANG J, ZHAO B, YANG C, et al. Rapeseed stand count estimation at leaf development stages with UAV imagery and convolutional neural networks[J]. Frontiers in Plant Science, 2020, 11: 617.
[16]
WU J, YANG G, YANG X, et al. Automatic counting of in situ rice seedlings from UAV images based on a deep fully convolutional neural network[J]. Remote Sensing, 2019, 11(6): 691.
[17]
XIONG H, CAO Z, LU H, et al. TasselNetv2: in-field counting of wheat spikes with context-augmented local regression networks[J]. Plant Methods, 2019, 15(1): 150.
[18]
饶雄飞, 周龙宇, 杨春雷, 等. 基于无人机多光谱影像和关键点检测的雪茄烟株数提取[J]. 农业机械学报, 2023, 54(3): 266-273.
RAO Xiongfei, ZHOU Longyu, YANG Chunlei, et al. Counting cigar tobacco plants from UAV multispectral images via key points detection approach[J], Transactions of the Chinese Society for Agricultural Machinery, 2023, 54(3): 266-273. (in Chinese)
[19]
FENG A, LI H, LIU Z, et al. Research on a rice counting algorithm based on an improved MCNN and a density map[J]. Entropy, 2021, 23(6): 721.
[20]
YANG H, YUAN H, WANG Y, et al. Rape seedling density estimation in-field conditions based on improved multi-column convolutional neural network[J]. Agronomy Journal, 2024, 116(3): 810-825.
[21]
CHEN Y, JIANG H, ZHANG X, et al. Refined feature fusion for in-field high-density and multi-scale rice panicle counting in UAV images[J]. Computers and Electronics in Agriculture, 2023, 211: 108032.
[22]
LAN S, REN Z, WU Y, et al. SaccadeNet: a fast and accurate object detector[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, 2020: 10397-10406.
[23]
WOO S, PARK J, LEE J Y, KWEON I S, Cbam: convolutional block attention module[C]//Proceedings of the European Conference on Computer Vision (ECCV), 2018: 3-19.
[24]
GAO J, WANG Q. SCAR: spatial-/channel-wise attention regression networks for crowd counting[J]. Neurocomputing, 2019, 363: 1-8.
2025年第56卷第12期
PDF下载
132
61
引用本文
BibTeX
文章信息
doi: 10.6041/j.issn.1000-1298.2025.12.051
  • 接收时间:2024-08-12
  • 首发时间:2026-09-30
  • 出版时间:2025-12-25
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-08-12
  • 修回日期:2024-10-14
基金
喜洋洋公司农业科研发展基金项目(112100000066)
作者信息
    1.中国科学技术大学自动化系,合肥 230026
    2.安徽喜洋洋农业科技有限公司,合肥 231521

通讯作者:

秦琳琳(1975—),女,高级工程师,博士,主要从事人工环境建模与控制研究,E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/nyjxxb/CN/10.6041/j.issn.1000-1298.2025.12.051
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