EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Structurally Consistent and Grounding-Aware Stagewise Reasoning for Referring Remote Sensing Image Segmentation

Shan Dong, Jianlin Xie, Liang Chen, He Chen, Baogui Qi, Yunqiu Ge

Remote Sensing · 28.03.2026 · 18 (7) · 1015

DOI: 10.3390/rs18071015 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 28.03.2026. Verfasst von Shan Dong und 5 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 72 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ye Qisun Science Foundation of the National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Referring Remote Sensing Image Segmentation (RRSIS) is a representative multimodal understanding task for remote sensing, which segments designated targets from remote images according to free-form natural language descriptions. However, complex remote sensing characteristics, such as cluttered backgrounds, large-scale variations, small scattered targets and repetitive textures, lead to unstable visual grounding and further spatial grounding drift, resulting in inaccurate segmentation results. Existing approaches typically perform implicit visual–linguistic fusion across encoding and decoding stages, entangling spatial grounding with mask refinement. This tightly coupled formulation lacks explicit structural constraints and is prone to cross-modal ambiguity, especially in complex remote sensing layouts. To address these limitations, we propose a Structurally consistent and Grounding-aware Stagewise Reasoning Framework (SGSRF) that follows a grounding-first, segmentation-second paradigm. The framework decomposes inference into three cascaded stages with progressively imposed structural constraints. First, Cross-modal Consistency Refinement (CCR) lays the foundation for stable spatial grounding by enhancing visual–textual structural alignment via CLIP-based features and Structural Consistency Regularization (SCR), producing well-aligned multimodal representations and reliable grounding cues. Second, Grounding-aware Prompt (GPG) Generation bridges grounding and segmentation by converting aligned representations into complementary sparse and dense prompts, which serve as explicit grounding guidance for the segmentation model. Third, Grounding Modulated Segmentation (GMS) leverages the Segment Anything Model (SAM) to generate fine-grained mask prediction under the joint guidance of prompts and grounding cues, improving spatial grounding stability and robustness to background interference and scale variation. Extensive experiments on three remote sensing benchmarks, namely RefSegRS, RRSIS-D, and RISBench, demonstrate that SGSRF achieves state-of-the-art performance. The proposed stagewise paradigm integrates structural alignment, explicit grounding, and prompt-driven segmentation into a unified framework, providing a practical and robust solution for RRSIS in real-world Earth observation applications.

Förderung

  • Ye Qisun Science Foundation of the National Natural Science Foundation of China

Zitieren

APA

Dong, S., Xie, J., Chen, L., Chen, H., Qi, B., & Ge, Y. (2026). Structurally Consistent and Grounding-Aware Stagewise Reasoning for Referring Remote Sensing Image Segmentation. Remote Sensing, 18(7), 1015. https://doi.org/10.3390/rs18071015

Vancouver

Dong S, Xie J, Chen L, Chen H, Qi B, Ge Y. Structurally Consistent and Grounding-Aware Stagewise Reasoning for Referring Remote Sensing Image Segmentation. Remote Sensing. 2026;18(7):1015. doi:10.3390/rs18071015

BibTeX
@article{Dong2026,
  author = {Dong, Shan and Xie, Jianlin and Chen, Liang and Chen, He and Qi, Baogui and Ge, Yunqiu},
  title = {Structurally Consistent and Grounding-Aware Stagewise Reasoning for Referring Remote Sensing Image Segmentation},
  journal = {Remote Sensing},
  year = {2026},
  volume = {18},
  number = {7},
  pages = {1015},
  doi = {10.3390/rs18071015}
}
Kostenlos starten →