View in EDS HTML Full Text PDF Full Text

Semantic Segmentation of Pegmatite Dikes in High-Resolution Remote Sensing Imagery Using GAD-UNet++ in the Yilanlike Area, South Tianshan.

Saved in:
Bibliographic Details
Title: Semantic Segmentation of Pegmatite Dikes in High-Resolution Remote Sensing Imagery Using GAD-UNet++ in the Yilanlike Area, South Tianshan.
Authors: Wu, Zirui1 (AUTHOR), Chen, Chuan1,2 (AUTHOR), Yu, Yuanjun3 (AUTHOR), Tian, Yong1,3 (AUTHOR), Yu, Jian1,2 (AUTHOR), Xia, Fang1,2,3 (AUTHOR) xf@xju.edu.cn
Source: Remote Sensing. Jun2026, Vol. 18 Issue 12, p1988. 36p.
Subjects: Remote sensing, Image segmentation, Pegmatites, Prospecting, Structural geology
Geographic Terms: Tien Shan, Xinjiang Uygur Zizhiqu (China)
Abstract: Highlights: What are the main findings? This study constructs a dedicated semantic segmentation dataset of pegmatite dikes for the Yilanlike area by using 0.66 m high-resolution RGB imagery for visual delineation and ZY1F hyperspectral data for spectral constraint and label refinement. The proposed GAD-UNet++ is designed as a pegmatite-dike-oriented semantic segmentation framework rather than a simple combination of existing modules. By integrating GhostNetV2, DFC, Coordinate Attention, and deep supervision within the UNet++ architecture, the model is adapted to address the slender morphology, discontinuous exposure, weak boundaries, background confusion, and class imbalance of pegmatite dikes in high-resolution remote sensing imagery. It achieved an mIoU of 93.11% and an F1-score of 94.95% on the test set, while reducing the number of parameters to 13.20 M and the inference time to 7.36 ms. Based on the segmentation results, 18 potential pegmatite dike enrichment zones were delineated in the study area. What are the implications of the main findings? The results indicate that combining lightweight feature extraction, long-range dependency modeling, and spatial attention enhancement is an effective strategy for segmenting slender geological targets with blurred boundaries and complex backgrounds, such as pegmatite dikes. The proposed dataset construction strategy and segmentation framework provide a practical technical pathway for pegmatite dike identification from high-resolution remote sensing imagery and support remote sensing-based rare-metal prospecting and regional geological interpretation. Pegmatite dikes are important prospecting indicators for rare-metal deposits, whereas traditional methods for pegmatite dike identification are constrained by the limited capability of human visual interpretation to capture information from remote sensing imagery, resulting in low identification accuracy and efficiency. In recent years, global research on semantic segmentation of different surface features and remote sensing-based mineral exploration using deep learning methods and high-resolution remote sensing imagery has made significant progress; however, studies on surface-exposed geological bodies such as pegmatite dikes remain highly insufficient. To address the key problem of efficiently identifying pegmatite dikes in remote sensing imagery, this study proposes an improved model based on UNet++, termed GAD-UNet++. In the field of remote sensing geology, this study constructed a pegmatite dike semantic segmentation dataset based on high-resolution RGB imagery by using 0.66 m RGB imagery for visual delineation and ZY1F hyperspectral data for spectral constraint and label refinement; on this basis, semantic segmentation of surface pegmatite dikes in the Yilanlike area of the South Tianshan Mountains, Xinjiang, was conducted using RGB remote sensing image patches as model input. Specifically, because pegmatite dikes are small targets characterized by slender structures, indistinct boundaries, and sparse regional distribution, this study introduced a lightweight feature extraction structure (GhostNetV2) and a long-range dependency attention module (DFC) at the encoder stage, and further incorporated the Coordinate Attention module (CA) to enhance spatial localization and boundary representation of the targets. Finally, focal cross-entropy loss and a deep supervision strategy were adopted to improve the accuracy of semantic information extraction for pegmatite dikes, as well as the training stability and segmentation accuracy under class-imbalance conditions. The results show that the proposed model achieved an mIoU of 93.11% and an F1-score of 94.95% on the test set. Compared with existing semantic segmentation models, the proposed model achieved superior performance in both identification accuracy and computational efficiency for pegmatite dikes. In addition, this study delineated 18 potential pegmatite dike enrichment zones in the Yilanlike area, providing technical support for remote sensing-based rare-metal prospecting and geological interpretation in the study area. [ABSTRACT FROM AUTHOR]
Copyright of Remote Sensing is the property of MDPI 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
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
Text:
  Availability: 1
Header DbId: egs
DbLabel: Engineering Source
An: 194915121
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Semantic Segmentation of Pegmatite Dikes in High-Resolution Remote Sensing Imagery Using GAD-UNet++ in the Yilanlike Area, South Tianshan.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Wu%2C+Zirui%22">Wu, Zirui</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Chen%2C+Chuan%22">Chen, Chuan</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Yu%2C+Yuanjun%22">Yu, Yuanjun</searchLink><relatesTo>3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Tian%2C+Yong%22">Tian, Yong</searchLink><relatesTo>1,3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Yu%2C+Jian%22">Yu, Jian</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Xia%2C+Fang%22">Xia, Fang</searchLink><relatesTo>1,2,3</relatesTo> (AUTHOR)<i> xf@xju.edu.cn</i>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Remote+Sensing%22">Remote Sensing</searchLink>. Jun2026, Vol. 18 Issue 12, p1988. 36p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Remote+sensing%22">Remote sensing</searchLink><br /><searchLink fieldCode="DE" term="%22Image+segmentation%22">Image segmentation</searchLink><br /><searchLink fieldCode="DE" term="%22Pegmatites%22">Pegmatites</searchLink><br /><searchLink fieldCode="DE" term="%22Prospecting%22">Prospecting</searchLink><br /><searchLink fieldCode="DE" term="%22Structural+geology%22">Structural geology</searchLink>
– Name: SubjectGeographic
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Tien+Shan%22">Tien Shan</searchLink><br /><searchLink fieldCode="DE" term="%22Xinjiang+Uygur+Zizhiqu+%28China%29%22">Xinjiang Uygur Zizhiqu (China)</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Highlights: What are the main findings? This study constructs a dedicated semantic segmentation dataset of pegmatite dikes for the Yilanlike area by using 0.66 m high-resolution RGB imagery for visual delineation and ZY1F hyperspectral data for spectral constraint and label refinement. The proposed GAD-UNet++ is designed as a pegmatite-dike-oriented semantic segmentation framework rather than a simple combination of existing modules. By integrating GhostNetV2, DFC, Coordinate Attention, and deep supervision within the UNet++ architecture, the model is adapted to address the slender morphology, discontinuous exposure, weak boundaries, background confusion, and class imbalance of pegmatite dikes in high-resolution remote sensing imagery. It achieved an mIoU of 93.11% and an F1-score of 94.95% on the test set, while reducing the number of parameters to 13.20 M and the inference time to 7.36 ms. Based on the segmentation results, 18 potential pegmatite dike enrichment zones were delineated in the study area. What are the implications of the main findings? The results indicate that combining lightweight feature extraction, long-range dependency modeling, and spatial attention enhancement is an effective strategy for segmenting slender geological targets with blurred boundaries and complex backgrounds, such as pegmatite dikes. The proposed dataset construction strategy and segmentation framework provide a practical technical pathway for pegmatite dike identification from high-resolution remote sensing imagery and support remote sensing-based rare-metal prospecting and regional geological interpretation. Pegmatite dikes are important prospecting indicators for rare-metal deposits, whereas traditional methods for pegmatite dike identification are constrained by the limited capability of human visual interpretation to capture information from remote sensing imagery, resulting in low identification accuracy and efficiency. In recent years, global research on semantic segmentation of different surface features and remote sensing-based mineral exploration using deep learning methods and high-resolution remote sensing imagery has made significant progress; however, studies on surface-exposed geological bodies such as pegmatite dikes remain highly insufficient. To address the key problem of efficiently identifying pegmatite dikes in remote sensing imagery, this study proposes an improved model based on UNet++, termed GAD-UNet++. In the field of remote sensing geology, this study constructed a pegmatite dike semantic segmentation dataset based on high-resolution RGB imagery by using 0.66 m RGB imagery for visual delineation and ZY1F hyperspectral data for spectral constraint and label refinement; on this basis, semantic segmentation of surface pegmatite dikes in the Yilanlike area of the South Tianshan Mountains, Xinjiang, was conducted using RGB remote sensing image patches as model input. Specifically, because pegmatite dikes are small targets characterized by slender structures, indistinct boundaries, and sparse regional distribution, this study introduced a lightweight feature extraction structure (GhostNetV2) and a long-range dependency attention module (DFC) at the encoder stage, and further incorporated the Coordinate Attention module (CA) to enhance spatial localization and boundary representation of the targets. Finally, focal cross-entropy loss and a deep supervision strategy were adopted to improve the accuracy of semantic information extraction for pegmatite dikes, as well as the training stability and segmentation accuracy under class-imbalance conditions. The results show that the proposed model achieved an mIoU of 93.11% and an F1-score of 94.95% on the test set. Compared with existing semantic segmentation models, the proposed model achieved superior performance in both identification accuracy and computational efficiency for pegmatite dikes. In addition, this study delineated 18 potential pegmatite dike enrichment zones in the Yilanlike area, providing technical support for remote sensing-based rare-metal prospecting and geological interpretation in the study area. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Remote Sensing is the property of MDPI 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=194915121
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.3390/rs18121988
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 36
        StartPage: 1988
    Subjects:
      – SubjectFull: Remote sensing
        Type: general
      – SubjectFull: Image segmentation
        Type: general
      – SubjectFull: Pegmatites
        Type: general
      – SubjectFull: Prospecting
        Type: general
      – SubjectFull: Structural geology
        Type: general
      – SubjectFull: Tien Shan
        Type: general
      – SubjectFull: Xinjiang Uygur Zizhiqu (China)
        Type: general
    Titles:
      – TitleFull: Semantic Segmentation of Pegmatite Dikes in High-Resolution Remote Sensing Imagery Using GAD-UNet++ in the Yilanlike Area, South Tianshan.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Wu, Zirui
      – PersonEntity:
          Name:
            NameFull: Chen, Chuan
      – PersonEntity:
          Name:
            NameFull: Yu, Yuanjun
      – PersonEntity:
          Name:
            NameFull: Tian, Yong
      – PersonEntity:
          Name:
            NameFull: Yu, Jian
      – PersonEntity:
          Name:
            NameFull: Xia, Fang
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 15
              M: 06
              Text: Jun2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 20724292
          Numbering:
            – Type: volume
              Value: 18
            – Type: issue
              Value: 12
          Titles:
            – TitleFull: Remote Sensing
              Type: main
ResultId 1