EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Semi-Supervised Structural Prior-Guided Network for Space Target Component Segmentation in ISAR Images

Yonghua He, Aoxiang Pan, Yonggang Li, Jiahao Wang, Wei Qu, Weigang Zhu, Wenhang Ji

Sensors · 27.07.2026 · 26 (15) · 4769

DOI: 10.3390/s26154769 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 27.07.2026. Verfasst von Yonghua He und 6 weiteren Autor:innen. Beteiligte Einrichtung: Space Engineering University, Beijing 101400, China. Die Arbeit nennt 60 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: State Administration of Science, Technology and Industry for National Defense, Defense Industrial Technology Development Program.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Segmenting key components of space targets using Inverse Synthetic Aperture Radar (ISAR) images is an important interpretation task in space situational awareness. However, the scarcity of pixel-level annotated data, inter-class confusion caused by morphological differences among multiple target classes, and the absence of structural priors for components restrict the performance improvement in existing deep models on this task. Therefore, this paper proposes a Semi-Supervised Structural Prior-Guided Network (SSPNet). First, a Gated Manifold-Constrained Hyper-Connections Vision Transformer (GMHC-ViT) encoder is proposed to broaden the feature representation space via parallel multi-feature streams with adaptive gating, thereby alleviating inter-class confusion and enhancing cross-category generalization. Second, a Prior-Guided Module (PGM) is proposed to extract shape and edge priors of components, and it adaptively enhances the weakly activated channels of encoder features through cross-attention, thereby injecting structural knowledge independent of image quality into the segmentation process. Furthermore, to effectively leverage large amounts of unlabeled data, a strong perturbation strategy tailored to the characteristics of ISAR images is designed for consistency regularization. Experimental results on a simulated ISAR dataset containing 38 classes of space targets demonstrate that SSPNet outperforms existing methods and exhibits strong segmentation capability even under low signal-to-noise ratio (SNR) conditions.

Schlagwörter

Stichwörter der Autor:innen

Inverse Synthetic Aperture Radar (ISAR)semantic segmentationsemi-supervised learningstructural prior

Autor:innen und Institutionen

  • Space Engineering University, Beijing 101400, China.

Förderung

  • State Administration of Science, Technology and Industry for National Defense
  • Defense Industrial Technology Development Program

Zitieren

APA

He, Y., Pan, A., Li, Y., Wang, J., Qu, W., Zhu, W., & Ji, W. (2026). Semi-Supervised Structural Prior-Guided Network for Space Target Component Segmentation in ISAR Images. Sensors, 26(15), 4769. https://doi.org/10.3390/s26154769

Vancouver

He Y, Pan A, Li Y, Wang J, Qu W, Zhu W, et al. Semi-Supervised Structural Prior-Guided Network for Space Target Component Segmentation in ISAR Images. Sensors. 2026;26(15):4769. doi:10.3390/s26154769

BibTeX
@article{He2026,
  author = {He, Yonghua and Pan, Aoxiang and Li, Yonggang and Wang, Jiahao and Qu, Wei and Zhu, Weigang and Ji, Wenhang},
  title = {Semi-Supervised Structural Prior-Guided Network for Space Target Component Segmentation in ISAR Images},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {15},
  pages = {4769},
  doi = {10.3390/s26154769}
}
Kostenlos starten →