航空学报 > 2026, Vol. 47 Issue (10): 532432-532432   doi: 10.7527/S1000-6893.2025.32432

航天遥感图像智能处理与分析专刊

自回归与反馈驱动的自适应矩形卷积全色锐化网络

段韶华1,2, 张淳杰1,2(), 刘传凯3,4, 郑晓龙5,6, 张济韬3,4   

  1. 1.北京交通大学 计算机科学与技术学院 信息科学研究所,北京 100044
    2.北京交通大学 计算机科学与技术学院 视觉智能交叉创新教育部国际合作联合实验室,北京 100044
    3.北京航天飞行控制中心,北京 100094
    4.航天飞行动力学技术重点实验室,北京 100094
    5.中国科学院 自动化研究所 多模态人工智能系统全国重点实验室,北京 100190
    6.中国科学院大学 人工智能学院,北京 100190
  • 收稿日期:2025-06-18 修回日期:2025-08-06 接受日期:2025-09-15 出版日期:2025-10-10 发布日期:2025-10-09
  • 通讯作者: 张淳杰 E-mail:cjzhang@bjtu.edu.cn
  • 基金资助:
    国家自然科学基金(62476021);国家自然科学基金(72225011);国家自然科学基金(72434005);国家自然科学基金(62373034);多模态人工智能系统全国重点实验室开放课题(MAIS2024106);中央高校基本科研业务费专项资金(2025JBZX062)

AFAR-Net: Autoregressive and feedback-driven adaptive rectangular convolution network for pansharpening

Shaohua DUAN1,2, Chunjie ZHANG1,2(), Chuankai LIU3,4, Xiaolong ZHENG5,6, Jitao ZHANG3,4   

  1. 1.Institute of Information Science,School of Computer Science and Technology,Beijing Jiaotong University,Beijing 100044,China
    2.Visual Intelligence +X International Cooperation Joint Laboratory of Ministry of Education,School of Computer Science and Technology,Beijing Jiaotong University,Beijing 100044,China
    3.Beijing Aerospace Control Center,Beijing 100094,China
    4.Key Laboratory of Science and Technology on Aerospace Flight Dynamics,Beijing 100094,China
    5.State Key Laboratory of Multimodal Artificial Intelligence Systems,Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China
    6.School of Artificial Intelligence,University of Chinese Academy of Sciences,Beijing 100190,China
  • Received:2025-06-18 Revised:2025-08-06 Accepted:2025-09-15 Online:2025-10-10 Published:2025-10-09
  • Contact: Chunjie ZHANG E-mail:cjzhang@bjtu.edu.cn
  • Supported by:
    National Natural Science Foundation of China(62476021);Open Project of State Key Laboratory of Multimodal Artificial Intelligence Systems(MAIS2024106);Fundamental Research Funds for the Central Universities(2025JBZX062)

摘要:

为提升遥感全色锐化任务中光谱信息的保真性和空间细节的恢复能力,提出一种基于编码-解码架构的深度全色锐化网络——自回归与反馈驱动的自适应矩形卷积网络(AFAR-Net)。该网络使用自回归机制,通过前一单元的预测结果优化当前单元,实现多级图像重建,其中反馈驱动融合模块被用于跨单元深度特征的高效融合,可有效提升光谱信息的一致性。同时,自适应卷积残差块可灵活调整卷积核规模与形状,增强网络对复杂空间结构的建模能力和恢复效果。通过设计轻量化融合头对多层级预测结果进行融合,从而提升最终重建结果的稳定性。在多个遥感数据集上的实验结果表明,所提网络在空间结构保持度、光谱角误差(SAM)和无参考图像质量指数(QNR)上均优于现有主流方法,展现出良好的泛化性和应用潜力。

关键词: 遥感图像, 全色锐化, 自回归, 自适应卷积, 深度学习

Abstract:

To enhance spectral fidelity and spatial detail restoration in remote sensing pansharpening tasks, this paper proposes a deep pansharpening network based on an encoder-decoder architecture, named the Autoregressive and Feedback-Driven Adaptive Rectangular Convolution Network for Pansharpening (AFAR-Net). The proposed network employs an autoregressive mechanism, where the output of the previous unit is used to optimize the current one, enabling progressive multi-level image reconstruction. Meanwhile, a feedback-driven fusion module is designed to efficiently integrate deep features across units, thereby enhancing spectral consistency. On the other hand, an adaptive convolutional residual block is introduced to flexibly adjust kernel sizes and shapes, strengthening the network’s ability to model and restore complex spatial structures. Finally, a lightweight fusion head is utilized to aggregate multi-level predictions, improving the stability of reconstruction. Experimental results on multiple remote sensing datasets demonstrate that the proposed network outperforms existing mainstream approaches in terms of spatial distortion, Spectral Angle Mapper (SAM), and the Quality with No Reference (QNR) index, showing strong generalization ability and application potential.

Key words: remote sensing image, pansharpening, autoregressive, adaptive convolution, deep learning

中图分类号: