航空学报 > 2025, Vol. 46 Issue (23): 632763-632763   doi: 10.7527/S1000-6893.2025.32763

干扰环境下无人机多源感知专栏

RS-AdaDiff:基于降质感知自适应估计的单步遥感图像超分辨率扩散模型

王飞1,2,3, 刘勇1,2,3, 姚嘉伟1,2,3, 朱轩磊1,2,3, 卢孝强4, 郭文星1,2,3, 张雪涛1,2,3, 郭宇1,2,3()   

  1. 1.西安交通大学 人机混合增强智能全国重点实验室,西安 710049
    2.西安交通大学 视觉信息与应用国家工程研究中心,西安 710049
    3.西安交通大学 人工智能与机器人研究所,西安 710049
    4.福州大学 物理与信息工程学院,福州 350108
  • 收稿日期:2025-09-06 修回日期:2025-09-24 接受日期:2025-10-20 出版日期:2025-11-20 发布日期:2025-11-13
  • 通讯作者: 郭宇 E-mail:yu.guo@xjtu.edu.cn
  • 基金资助:
    国家重大科技专项(2009XJTU0016)

RS-AdaDiff: One-step remote sensing image super-resolution diffusion model with degradation-aware adaptive estimation

Fei WANG1,2,3, Yong LIU1,2,3, Jiawei YAO1,2,3, Xuanlei ZHU1,2,3, Xiaoqiang LU4, Wenxing GUO1,2,3, Xuetao ZHANG1,2,3, Yu GUO1,2,3()   

  1. 1.National Key Laboratory of Human-Machine Hybrid Augmented Intelligence,Xi’an Jiaotong University,Xi’an 710049,China
    2.National Engineering Research Center of Visual Information and Applications,Xi’an Jiaotong University,Xi’an 710049,China
    3.Institute of Artificial Intelligence and Robotics,Xi’an Jiaotong University,Xi’an 710049,China
    4.College of Physics and Information Engineering,Fuzhou University,Fuzhou 350108,China
  • Received:2025-09-06 Revised:2025-09-24 Accepted:2025-10-20 Online:2025-11-20 Published:2025-11-13
  • Contact: Yu GUO E-mail:yu.guo@xjtu.edu.cn
  • Supported by:
    National Major Science and Technology Projects of China(2009XJTU0016)

摘要:

扩散模型在生成逼真图像细节方面展现出巨大潜力。然而,现有的扩散模型主要基于自然图像进行训练,将这些模型应用于遥感图像超分辨率任务仍然面临巨大挑战。此外,这些模型在推理时需要数十或上百次的迭代采样,导致计算成本高昂,并限制了它们在实际应用中的适用性。为此,提出一种基于降质感知自适应估计的单步遥感图像超分辨率扩散模型(RS-AdaDiff),兼顾重建性能与推理效率。具体而言,提出了一个基于降质感知的时间步估计模块,可通过估计输入图像退化程度的扩散模型自适应估计扩散时间步,从而将迭代去噪过程重构为从低分辨率到高分辨率图像的单步重建过程,大幅加快推理速度。同时,将可训练的轻量LoRA网络层集成到预先训练的扩散模型中,并利用遥感图像数据集对其进行微调,以消除数据分布差异造成的领域差距问题。此外,为了充分利用预训练模型的图像先验,引入了分布对比匹配蒸馏。通过KL散度正则化,使重建的超分图像在特征空间中更接近高分辨率图像并远离低分辨率图像,从而提升生成质量。最后,还提出特征-边缘联合感知相似度损失,以增强结构信息的感知能力,改善边缘模糊和纹理失真问题。大量实验结果表明:提出的RS-AdaDiff在多个公开遥感数据集上均优于现有先进方法,在定量指标和视觉质量方面均取得显著提升,能够生成结构清晰、细节丰富的超分辨率遥感图像。

关键词: 遥感图像超分辨率, 扩散模型, 自适应估计, 计算机视觉, 航空航天

Abstract:

Diffusion models have demonstrated great potential in generating realistic image details. However, existing diffusion models are primarily trained on natural images, making their application to remote sensing image super-resolution highly challenging. Moreover, these models typically require dozens or even hundreds of iterative sampling steps during inference, resulting in high computational costs and limited practicality. To address these issues, this paper proposes a degradation-aware adaptive estimation-based single-step remote sensing image super-resolution diffusion model (RS-AdaDiff), which balances reconstruction performance and inference efficiency. Specifically, we propose a degradation-aware timestep estimation module that adaptively estimates the diffusion timestep for the diffusion model by assessing the degradation level of the input image. This approach reconstructs the iterative denoising process into a single-step reconstruction from low-resolution to high-resolution images, thereby significantly accelerating inference. Meanwhile, we integrate trainable lightweight LoRA layers into a pre-trained diffusion model and fine-tune it on a remote sensing image dataset to mitigate the domain gap caused by data distribution differences. Additionally, to fully leverage the image priors of the pre-trained model, we introduce distribution contrastive matching distillation. By regularizing the KL divergence, the reconstructed super-resolved images are brought closer to high-resolution images and farther from low-resolution images in the feature space, thereby improving generation quality. Finally, we propose a feature-edge joint perceptual similarity loss to enhance the perception of structural information and mitigate issues such as edge blur and texture distortion. Extensive experimental results demonstrate that the proposed RS-AdaDiff outperforms existing state-of-the-art methods on multiple public remote sensing datasets, achieving significant improvements in both quantitative metrics and visual quality, and producing super-resolved remote sensing images with clearer structures and richer details.

Key words: remote sensing image super-resolution, diffusion model, adaptive estimation, computer vision, aerospace

中图分类号: