View in EDS PDF Full Text

Multi-source feature fusion network for grape berry instance segmentation.

Saved in:
Bibliographic Details
Title: Multi-source feature fusion network for grape berry instance segmentation.
Authors: Yao, Mingcheng1 yaomingcheng@qq.com, Yang, Xiaoxia1 yangxx@sdau.edu.cn, Zhang, Chengming1 fanjl@bnu.edu.cn, Wu, Menxin2 hongyinglhy@126.com, Fan, Jinlong3 weijian0269@163.com, Li, Feng4 19853838025@163.com, Li, Hongying5 15621888796@163.com, Wei, Jianguo6 13623933329@163.com, Liu, Shujun1, Zhang, Huake1 chming@sdau.edu.cn, Jin, Yanhui1 lfeng1029@163.com
Source: International Journal of Agricultural & Biological Engineering. Apr2026, Vol. 19 Issue 2, p294-302. 9p.
Subjects: Image segmentation, Edge detection (Image processing), Morphology, Transformer models, Phenotypes
Abstract: Accurate delineation of grape berry boundaries is essential for phenotypic measurement and growth assessment. This study proposes a multi-source feature fusion network (MFFNet) for instance segmentation of grape berries in dense clusters with frequent overlaps and blurred edges. MFFNet employs two parallel branches for feature extraction: a Swin Transformer backbone to capture hierarchical semantic features and an edge-detection branch that predicts an edge probability map to provide boundary cues. To address the substantial scale variation within a single image, the multilevel semantic features were enhanced using Adaptive Spatial Feature Fusion (ASFF). The edge probability map was introduced twice into the ASFF-enhanced multi-scale features. First, edge cues were injected into the highest-resolution fused feature map to strengthen global boundary awareness across the cluster. Second, during mask generation, edge cues were reintroduced within each candidate instance region to refine local contours and improve the separation in the adhered areas. Experiments on a custom dataset collected in Yinchuan, Ningxia, showed that MFFNet achieved an mAP5b0ox of 93.4% and mAP5m0ask of 93.4%, outperforming representative baselines, including Mask2Former and HTC. The proposed model remained stable on images with severe berry overlap and indistinct edges, supporting practical grape growth monitoring. [ABSTRACT FROM AUTHOR]
Copyright of International Journal of Agricultural & Biological Engineering is the property of International Journal of Agricultural & Biological Engineering and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
FullText Links:
  – Type: pdflink
Text:
  Availability: 0
Header DbId: egs
DbLabel: Engineering Source
An: 194079842
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Multi-source feature fusion network for grape berry instance segmentation.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Yao%2C+Mingcheng%22">Yao, Mingcheng</searchLink><relatesTo>1</relatesTo><i> yaomingcheng@qq.com</i><br /><searchLink fieldCode="AR" term="%22Yang%2C+Xiaoxia%22">Yang, Xiaoxia</searchLink><relatesTo>1</relatesTo><i> yangxx@sdau.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Zhang%2C+Chengming%22">Zhang, Chengming</searchLink><relatesTo>1</relatesTo><i> fanjl@bnu.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Wu%2C+Menxin%22">Wu, Menxin</searchLink><relatesTo>2</relatesTo><i> hongyinglhy@126.com</i><br /><searchLink fieldCode="AR" term="%22Fan%2C+Jinlong%22">Fan, Jinlong</searchLink><relatesTo>3</relatesTo><i> weijian0269@163.com</i><br /><searchLink fieldCode="AR" term="%22Li%2C+Feng%22">Li, Feng</searchLink><relatesTo>4</relatesTo><i> 19853838025@163.com</i><br /><searchLink fieldCode="AR" term="%22Li%2C+Hongying%22">Li, Hongying</searchLink><relatesTo>5</relatesTo><i> 15621888796@163.com</i><br /><searchLink fieldCode="AR" term="%22Wei%2C+Jianguo%22">Wei, Jianguo</searchLink><relatesTo>6</relatesTo><i> 13623933329@163.com</i><br /><searchLink fieldCode="AR" term="%22Liu%2C+Shujun%22">Liu, Shujun</searchLink><relatesTo>1</relatesTo><br /><searchLink fieldCode="AR" term="%22Zhang%2C+Huake%22">Zhang, Huake</searchLink><relatesTo>1</relatesTo><i> chming@sdau.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Jin%2C+Yanhui%22">Jin, Yanhui</searchLink><relatesTo>1</relatesTo><i> lfeng1029@163.com</i>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22International+Journal+of+Agricultural+%26+Biological+Engineering%22">International Journal of Agricultural & Biological Engineering</searchLink>. Apr2026, Vol. 19 Issue 2, p294-302. 9p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Image+segmentation%22">Image segmentation</searchLink><br /><searchLink fieldCode="DE" term="%22Edge+detection+%28Image+processing%29%22">Edge detection (Image processing)</searchLink><br /><searchLink fieldCode="DE" term="%22Morphology%22">Morphology</searchLink><br /><searchLink fieldCode="DE" term="%22Transformer+models%22">Transformer models</searchLink><br /><searchLink fieldCode="DE" term="%22Phenotypes%22">Phenotypes</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Accurate delineation of grape berry boundaries is essential for phenotypic measurement and growth assessment. This study proposes a multi-source feature fusion network (MFFNet) for instance segmentation of grape berries in dense clusters with frequent overlaps and blurred edges. MFFNet employs two parallel branches for feature extraction: a Swin Transformer backbone to capture hierarchical semantic features and an edge-detection branch that predicts an edge probability map to provide boundary cues. To address the substantial scale variation within a single image, the multilevel semantic features were enhanced using Adaptive Spatial Feature Fusion (ASFF). The edge probability map was introduced twice into the ASFF-enhanced multi-scale features. First, edge cues were injected into the highest-resolution fused feature map to strengthen global boundary awareness across the cluster. Second, during mask generation, edge cues were reintroduced within each candidate instance region to refine local contours and improve the separation in the adhered areas. Experiments on a custom dataset collected in Yinchuan, Ningxia, showed that MFFNet achieved an mAP5b0ox of 93.4% and mAP5m0ask of 93.4%, outperforming representative baselines, including Mask2Former and HTC. The proposed model remained stable on images with severe berry overlap and indistinct edges, supporting practical grape growth monitoring. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of International Journal of Agricultural & Biological Engineering is the property of International Journal of Agricultural & Biological Engineering and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=194079842
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.25165/j.ijabe.20261902.10132
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 9
        StartPage: 294
    Subjects:
      – SubjectFull: Image segmentation
        Type: general
      – SubjectFull: Edge detection (Image processing)
        Type: general
      – SubjectFull: Morphology
        Type: general
      – SubjectFull: Transformer models
        Type: general
      – SubjectFull: Phenotypes
        Type: general
    Titles:
      – TitleFull: Multi-source feature fusion network for grape berry instance segmentation.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Yao, Mingcheng
      – PersonEntity:
          Name:
            NameFull: Yang, Xiaoxia
      – PersonEntity:
          Name:
            NameFull: Zhang, Chengming
      – PersonEntity:
          Name:
            NameFull: Wu, Menxin
      – PersonEntity:
          Name:
            NameFull: Fan, Jinlong
      – PersonEntity:
          Name:
            NameFull: Li, Feng
      – PersonEntity:
          Name:
            NameFull: Li, Hongying
      – PersonEntity:
          Name:
            NameFull: Wei, Jianguo
      – PersonEntity:
          Name:
            NameFull: Liu, Shujun
      – PersonEntity:
          Name:
            NameFull: Zhang, Huake
      – PersonEntity:
          Name:
            NameFull: Jin, Yanhui
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 04
              Text: Apr2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 19346344
          Numbering:
            – Type: volume
              Value: 19
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: International Journal of Agricultural & Biological Engineering
              Type: main
ResultId 1