融合Transformer与原型自监督的苹果叶部病害识别
TP391.4%S431.9; 为了缓解苹果叶部病害(apple leaf diseases,ALD)识别存在"类内差异大、类间差异小"的问题,该研究设计了一种融合 Transformer 与原型自监督(fusion transformer and prototype self-supervised,FTPSS)的模型,以进一步提高ALD识别精度.首先,利用Resnet50作为骨干网络,以提取ALD图像的多级特征图谱;然后,构造了一个简化自注意力(simplified self-attention,SSA)机制,且将其与空间注意力引导可变形卷积(spatial attent...
Saved in:
| Published in | 农业工程学报 Vol. 40; no. 23; pp. 208 - 216 |
|---|---|
| Main Authors | , , |
| Format | Journal Article |
| Language | Chinese |
| Published |
西安市公共安全图像处理技术及应用重点实验室西安 710121%西安邮电大学通信与信息工程学院,西安 710121
01.12.2024
西安邮电大学通信与信息工程学院,西安 710121 |
| Subjects | |
| Online Access | Get full text |
| ISSN | 1002-6819 |
| DOI | 10.11975/j.issn.1002-6819.202405187 |
Cover
| Abstract | TP391.4%S431.9; 为了缓解苹果叶部病害(apple leaf diseases,ALD)识别存在"类内差异大、类间差异小"的问题,该研究设计了一种融合 Transformer 与原型自监督(fusion transformer and prototype self-supervised,FTPSS)的模型,以进一步提高ALD识别精度.首先,利用Resnet50作为骨干网络,以提取ALD图像的多级特征图谱;然后,构造了一个简化自注意力(simplified self-attention,SSA)机制,且将其与空间注意力引导可变形卷积(spatial attention guided deformable convolution,SAG-DC)相结合,设计了一种简化自注意力可变卷积 Transformer(simplified self-attention and deformable convolution transformer,SSADC-TF)编码器,用于对主干网络提取的多级特征图谱进行交互融合,以增强模型对ALD图像中不规则病变区域的感知能力;最后,构造了一个原型自监督(prototype self-supervised,PSS)学习模块,即通过构造"正交"与"聚集"二个自监督损失函数用于约束模型的训练,以缓解ALD图像识别中存在的语义鸿沟问题.基于标准图像集与真实图像集的对比试验结果表明,ALD图像经SSADC-TF逐层特征融合与PSS学习之后,FTPSS模型识别精度分别达到98.61％与98.73％,较基线模型分别提高5.15与4.49个百分点,能够满足智慧农业ALD识别的应用需求. |
|---|---|
| AbstractList | TP391.4%S431.9; 为了缓解苹果叶部病害(apple leaf diseases,ALD)识别存在"类内差异大、类间差异小"的问题,该研究设计了一种融合 Transformer 与原型自监督(fusion transformer and prototype self-supervised,FTPSS)的模型,以进一步提高ALD识别精度.首先,利用Resnet50作为骨干网络,以提取ALD图像的多级特征图谱;然后,构造了一个简化自注意力(simplified self-attention,SSA)机制,且将其与空间注意力引导可变形卷积(spatial attention guided deformable convolution,SAG-DC)相结合,设计了一种简化自注意力可变卷积 Transformer(simplified self-attention and deformable convolution transformer,SSADC-TF)编码器,用于对主干网络提取的多级特征图谱进行交互融合,以增强模型对ALD图像中不规则病变区域的感知能力;最后,构造了一个原型自监督(prototype self-supervised,PSS)学习模块,即通过构造"正交"与"聚集"二个自监督损失函数用于约束模型的训练,以缓解ALD图像识别中存在的语义鸿沟问题.基于标准图像集与真实图像集的对比试验结果表明,ALD图像经SSADC-TF逐层特征融合与PSS学习之后,FTPSS模型识别精度分别达到98.61％与98.73％,较基线模型分别提高5.15与4.49个百分点,能够满足智慧农业ALD识别的应用需求. |
| Abstract_FL | Apple leaf diseases(ALD)identification can be characterized by"significant intra-class variation and subtle inter-class differences".In this study,an innovative model was presented to integrate transformer with prototype self-supervised(FTPSS)learning.This improved model aimed to significantly elevate the precision of ALD recognition,thereby enhancing disease management strategies in orchards.The ResNet50 was utilized as the backbone network in the FTPSS model.This robust architecture was employed to extract multi-level feature maps from ALD images,in order to capture the intricate details for accurate disease identification.An encoder design was also integrated a simplified self-attention(SSA)mechanism with spatial attention guided deformable convolution(SAG-DC).The simplified self-attention and deformable convolution transformer(SSADC-TF)was used to facilitate the effective interaction and fusion of multi-level feature maps.The extracted features were then processed.The sensitivity of model was enhanced for the irregular lesion areas within ALD images.SSADC-TF was significantly distinguished among different disease manifestations.A prototype self-supervised(PSS)learning module was introduced to further verify the performance of model.Two self-supervised loss functions:"Orthogonality"and"Clustering"were selected in the module.In the"Orthogonality"loss,the feature representations of different ALD classes were orthogonal to each other.A clear separation among classes was promoted to enhance the identification of the model.Meanwhile,the"Clustering"loss was used to tighten the intra-class compactness,thus ensuring that the variations within the same class was suitable for the robustness of the model.Extensive experiments were conducted on both standard and real-world image datasets,indicating the remarkable effectiveness of FTPSS model.The FTPSS model was achieved in a recognition accuracy of 98.61％on the standard image set,indicating a significant improvement of 5.15 percentage points over the baseline model.Similarly,the FTPSS model was obtained an accuracy of 98.73％on the real-world image set,indicating an enhancement of 4.49 percentage points,compared with the baseline.These results underscored the robust performance of FTPSS model to identify ALD,even in the presence of significant intra-class variation and subtle inter-class differences.The FTPSS model was attributed to the innovative integration of Transformer with Prototype Self-Supervised learning.There were the powerful feature extraction of ResNet50.SSADC-TF was also enhanced feature interaction and fusion.The complex details in ALD images were captured to achieve in a 2.40 percentage point improvement.Furthermore,the PSS learning module was introduced to mitigate the semantic gap,where the model was generalized well to new,unseen ALD cases.The accuracy of ALD image recognition increased by 2.69 percentage points.In conclusion,the FTPSS model shared a significant advancement in ALD recognition,with the potential to revolutionize disease management strategies in orchards.The precise,timely information can be expected to apply into the automatic process of disease detection ALD,thereby preserving the health and productivity of the orchards.This finding can greatly contribute to the field of precision agriculture using advanced deep learning techniques. |
| Author | 李大湘 刘颖 张雯凯 |
| AuthorAffiliation | 西安邮电大学通信与信息工程学院,西安 710121;西安市公共安全图像处理技术及应用重点实验室西安 710121%西安邮电大学通信与信息工程学院,西安 710121 |
| AuthorAffiliation_xml | – name: 西安邮电大学通信与信息工程学院,西安 710121;西安市公共安全图像处理技术及应用重点实验室西安 710121%西安邮电大学通信与信息工程学院,西安 710121 |
| Author_FL | LIU Ying LI Daxiang ZHANG Wenkai |
| Author_FL_xml | – sequence: 1 fullname: LI Daxiang – sequence: 2 fullname: ZHANG Wenkai – sequence: 3 fullname: LIU Ying |
| Author_xml | – sequence: 1 fullname: 李大湘 – sequence: 2 fullname: 张雯凯 – sequence: 3 fullname: 刘颖 |
| BookMark | eNo9jz1Lw0Ach2-oYK39FA5Oif-7S3LJIkjxDQoudS6Xy11p0QvkEHX2rVBadRJ8iYOLKOggiH4fz8RvYURx-sEz_B6eGVTTqZYIzWFwMY6YvzBw-8ZoFwMQJwhx5BIgHvg4ZDVU_6fTqGlMP644ZQAerqPFMh_bs2En49qoNNuW2cfbxE5ubT4qTx6Kq_Pi5q64PCxH75_5tT19_Tq4Ly6O7NNL-Xxsh4-zaErxLSObf9tAmyvLndaa095YXW8ttR2DgWCH-hwEY4pSHnqJ5AIiLJNQKRkoISgAYC5CFUeBiglESexhnwlGBA2TWAYBbaD5399drhXXve4g3cl0Zezq_Z7Yi39qCa1U9BszsWGV |
| ClassificationCodes | TP391.4%S431.9 |
| ContentType | Journal Article |
| Copyright | Copyright © Wanfang Data Co. Ltd. All Rights Reserved. |
| Copyright_xml | – notice: Copyright © Wanfang Data Co. Ltd. All Rights Reserved. |
| DBID | 2B. 4A8 92I 93N PSX TCJ |
| DOI | 10.11975/j.issn.1002-6819.202405187 |
| DatabaseName | Wanfang Data Journals - Hong Kong WANFANG Data Centre Wanfang Data Journals 万方数据期刊 - 香港版 China Online Journals (COJ) China Online Journals (COJ) |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Agriculture |
| DocumentTitle_FL | Identifying apple leaf diseases by integrating transformer and prototype self-supervised learning |
| EndPage | 216 |
| ExternalDocumentID | nygcxb202423021 |
| GrantInformation_xml | – fundername: (国家自然科学基金); (陕西省自然科学基金) funderid: (国家自然科学基金); (陕西省自然科学基金) |
| GroupedDBID | -04 2B. 4A8 5XA 5XE 92G 92I 93N ABDBF ABJNI ACGFO ACGFS ACUHS AEGXH AIAGR ALMA_UNASSIGNED_HOLDINGS CCEZO CHDYS CW9 EOJEC FIJ IPNFZ OBODZ PSX RIG TCJ TGD TUS U1G U5N |
| ID | FETCH-LOGICAL-s1021-35a0c77f33a84deac091ed8ffe6fcc30001ac8fb96fb209db4157c72c38dbe663 |
| ISSN | 1002-6819 |
| IngestDate | Thu May 29 04:08:37 EDT 2025 |
| IsPeerReviewed | false |
| IsScholarly | true |
| Issue | 23 |
| Keywords | 病害 自监督学习 apple 苹果 计算机视觉 computer vision Transformer编码器 diseases self-supervised learning transformer encoder |
| Language | Chinese |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-s1021-35a0c77f33a84deac091ed8ffe6fcc30001ac8fb96fb209db4157c72c38dbe663 |
| PageCount | 9 |
| ParticipantIDs | wanfang_journals_nygcxb202423021 |
| PublicationCentury | 2000 |
| PublicationDate | 2024-12-01 |
| PublicationDateYYYYMMDD | 2024-12-01 |
| PublicationDate_xml | – month: 12 year: 2024 text: 2024-12-01 day: 01 |
| PublicationDecade | 2020 |
| PublicationTitle | 农业工程学报 |
| PublicationTitle_FL | Transactions of the Chinese Society of Agricultural Engineering |
| PublicationYear | 2024 |
| Publisher | 西安市公共安全图像处理技术及应用重点实验室西安 710121%西安邮电大学通信与信息工程学院,西安 710121 西安邮电大学通信与信息工程学院,西安 710121 |
| Publisher_xml | – name: 西安邮电大学通信与信息工程学院,西安 710121 – name: 西安市公共安全图像处理技术及应用重点实验室西安 710121%西安邮电大学通信与信息工程学院,西安 710121 |
| SSID | ssib051370041 ssj0041925 ssib001101065 ssib023167668 |
| Score | 2.4840825 |
| Snippet | TP391.4%S431.9; 为了缓解苹果叶部病害(apple leaf diseases,ALD)识别存在"类内差异大、类间差异小"的问题,该研究设计了一种融合 Transformer 与原型自监督(fusion... |
| SourceID | wanfang |
| SourceType | Aggregation Database |
| StartPage | 208 |
| Title | 融合Transformer与原型自监督的苹果叶部病害识别 |
| URI | https://d.wanfangdata.com.cn/periodical/nygcxb202423021 |
| Volume | 40 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVEBS databaseName: EBSCOhost Academic Search Ultimate issn: 1002-6819 databaseCode: ABDBF dateStart: 20140101 customDbUrl: https://search.ebscohost.com/login.aspx?authtype=ip,shib&custid=s3936755&profile=ehost&defaultdb=asn isFulltext: true dateEnd: 99991231 titleUrlDefault: https://search.ebscohost.com/direct.asp?db=asn omitProxy: true ssIdentifier: ssj0041925 providerName: EBSCOhost |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwnV1JaxRBFC7CBEQP4oo7AX0nmdh7V12Eqp5ugqB4SCC3MNXTHU8TyAKaq1sgGPUkuMSDF1HQgyD6fxxn_Be-V10z09lwuTTFq9dvqa-7Xr3mVTVjV3RYxrnw3KavvbgZRFw3RVB6Ta1dx-mUbtnW9EH_5q1oZi64MR_OTzRu16qW1lb1dL6-776S_0EVaYgr7ZL9B2RHQpGAbcQXr4gwXv8KY0g5iBR4C9IQhAOczw7XoYhEGoDiwFPqxKvIDBeyK7qPxyAlpDEIBcI1jRZI3zQk8MDwKFAC0ojuEomRk4GKIBXAfZDcMMfAQ-qSKSif7pIZ8MgwY1vVV7-GGBlRxjZUhBSFloQkShqNJAotiUgvl9g1fCaMIS3rkAxAxkRBAwUfs6C4BKRDJqJj0vhMrmZ1FjRMcGKRHlR_uh1--PCCXUUkxiEUmlkXuTA60FLPSEJiYhvK3cFDXdwMOY5iBYJP41cZTwMcG8yGjnLHOJgMjc6ISLokiIo5MAIF-UOQx2SDElYpQkQOoZzEUmSw1_irMZ20htEr3M8xQRKxXSlTYW2gR6AIUxsjDYIZSLf2mI0o6I9r3TgQXMTHjL-XHGhlLWBSRI24DXs2olYHcNmZw_Pr8dHhtaWWV22z3RvFRRyaME4qpkcqpukpwCBiV2g7j0nv3lvM72riwKSaTpaY9DDOOw02KVVLZeMUwaWvIKMY5tFJENE45Q5dn374MCoToyKJ0FRMWCMOsctDE68dbKDZAtgt293F2mp19hg7atPMKVnNGcfZxPqdE-yIXFy2R-0UJ9n1wfaT3rON2ozx49tWb-ttb3tz8PhD_9Xz_pt3_ZcPBpvff26_7j39-uv--_6Lh71PXwafH_U2Pp5ic1k6m8w07d9Umisu1WH5YdvJ47j0_TYPOrjewkyh6PCyLKIyz31K9to5L7WIShxG0dG4tI_z2Mt93tEFJianWaO71C3OsCm_g2lBnuehaFNFSYRzfYhpTqQLT-fa5WfZlPV-wc6WKwu78Dn3Z5bz7PD4tb_AGqvLa8VFzABW9SUL6m_2Prp2 |
| linkProvider | EBSCOhost |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=%E8%9E%8D%E5%90%88Transformer%E4%B8%8E%E5%8E%9F%E5%9E%8B%E8%87%AA%E7%9B%91%E7%9D%A3%E7%9A%84%E8%8B%B9%E6%9E%9C%E5%8F%B6%E9%83%A8%E7%97%85%E5%AE%B3%E8%AF%86%E5%88%AB&rft.jtitle=%E5%86%9C%E4%B8%9A%E5%B7%A5%E7%A8%8B%E5%AD%A6%E6%8A%A5&rft.au=%E6%9D%8E%E5%A4%A7%E6%B9%98&rft.au=%E5%BC%A0%E9%9B%AF%E5%87%AF&rft.au=%E5%88%98%E9%A2%96&rft.date=2024-12-01&rft.pub=%E8%A5%BF%E5%AE%89%E5%B8%82%E5%85%AC%E5%85%B1%E5%AE%89%E5%85%A8%E5%9B%BE%E5%83%8F%E5%A4%84%E7%90%86%E6%8A%80%E6%9C%AF%E5%8F%8A%E5%BA%94%E7%94%A8%E9%87%8D%E7%82%B9%E5%AE%9E%E9%AA%8C%E5%AE%A4%E8%A5%BF%E5%AE%89+710121%25%E8%A5%BF%E5%AE%89%E9%82%AE%E7%94%B5%E5%A4%A7%E5%AD%A6%E9%80%9A%E4%BF%A1%E4%B8%8E%E4%BF%A1%E6%81%AF%E5%B7%A5%E7%A8%8B%E5%AD%A6%E9%99%A2%2C%E8%A5%BF%E5%AE%89+710121&rft.issn=1002-6819&rft.volume=40&rft.issue=23&rft.spage=208&rft.epage=216&rft_id=info:doi/10.11975%2Fj.issn.1002-6819.202405187&rft.externalDocID=nygcxb202423021 |
| thumbnail_s | http://utb.summon.serialssolutions.com/2.0.0/image/custom?url=http%3A%2F%2Fwww.wanfangdata.com.cn%2Fimages%2FPeriodicalImages%2Fnygcxb%2Fnygcxb.jpg |