Salient Region-Guided Spacecraft Image Arbitrary-Scale Super-Resolution Network

2025-04-25

AI Summary

Spacecraft image super-resolution (SR) suffers from significant feature disparity between the spacecraft’s core region and the deep-space background, leading to background noise amplification. Method: This paper proposes a saliency-region-guided arbitrary-scale SR network. It introduces— for the first time in spacecraft SR—an implicit modulation mechanism guided by core-region saliency detection, coupled with an Adaptive Weighted Feature Fusion Enhancement Module (AFFEM) that dynamically strengthens feature responses in the core region. Technical implementation integrates a pre-trained saliency detection model and a Spacecraft Core Region Recognition Block (SCRRB). Contribution/Results: Evaluated on multi-scale spacecraft imagery, the method outperforms state-of-the-art (SOTA) approaches, effectively suppressing black-sky background noise while significantly improving reconstruction fidelity of structural details in the core region. It establishes a new high-fidelity SR paradigm for remote sensing and space object identification.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Spacecraft image super-resolution seeks to enhance low-resolution spacecraft images into high-resolution ones. Although existing arbitrary-scale super-resolution methods perform well on general images, they tend to overlook the difference in features between the spacecraft core region and the large black space background, introducing irrelevant noise. In this paper, we propose a salient region-guided spacecraft image arbitrary-scale super-resolution network (SGSASR), which uses features from the spacecraft core salient regions to guide latent modulation and achieve arbitrary-scale super-resolution. Specifically, we design a spacecraft core region recognition block (SCRRB) that identifies the core salient regions in spacecraft images using a pre-trained saliency detection model. Furthermore, we present an adaptive-weighted feature fusion enhancement mechanism (AFFEM) to selectively aggregate the spacecraft core region features with general image features by dynamic weight parameter to enhance the response of the core salient regions. Experimental results demonstrate that the proposed SGSASR outperforms state-of-the-art approaches.

Problem

• Enhancing low-resolution spacecraft images to high-resolution
• Reducing irrelevant noise in spacecraft core region features
• Achieving arbitrary-scale super-resolution with salient region guidance

Innovation

• Salient region-guided arbitrary-scale super-resolution network
• Spacecraft core region recognition block (SCRRB)
• Adaptive-weighted feature fusion enhancement mechanism (AFFEM)