航空学报 > 2025, Vol. 46 Issue (11): 531343-531343   doi: 10.7527/S1000-6893.2024.31343

双光载荷图像融合及其在低空遥感中的应用

孙彬1,2,3(), 游航1,2,3, 李文博1,2,3, 刘祥瑞1,2,3, 马佳义4   

  1. 1.电子科技大学 航空航天学院,成都 611731
    2.自适应光学全国重点实验室,成都 610209
    3.飞行器集群智能感知与协同控制四川省重点实验室,成都 611731 4.武汉大学 电子信息学院,武汉 430072
  • 收稿日期:2024-10-08 修回日期:2024-11-20 接受日期:2024-12-06 出版日期:2024-12-30 发布日期:2024-12-23
  • 通讯作者: 孙彬 E-mail:sunbinhust@uestc.edu.cn
  • 基金资助:
    国家自然科学基金(U23B2050);四川省科技计划资助(2022YFG0050);中央高校基本科研业务费专项资金(ZYGX2020ZB032)

Dual-band payload image fusion and its applications in low-altitude remote sensing

Bin SUN1,2,3(), Hang YOU1,2,3, Wenbo LI1,2,3, Xiangrui LIU1,2,3, Jiayi MA4   

  1. 1.School of Aeronautics and Astronautics,University of Electronic Science and Technology of China,Chengdu 611731,China
    2.National Laboratory on Adaptive Optics,Chengdu 610209,China
    3.Aircraft Swarm Intelligent Sensing and Cooperative Control Key Laboratory of Sichuan Province,Chengdu 611731,China
    4.Electronic Information School,Wuhan University,Wuhan 430072,China
  • Received:2024-10-08 Revised:2024-11-20 Accepted:2024-12-06 Online:2024-12-30 Published:2024-12-23
  • Contact: Bin SUN E-mail:sunbinhust@uestc.edu.cn
  • Supported by:
    National Natural Science Foundation of China(U23B2050);Sichuan Science and Technology Program(2022YFG0050);Fundamental Research Funds for the Central Universities(ZYGX2020ZB032)

摘要:

双光载荷图像融合在目标监测、灾情预警及专业巡检等低空遥感领域有着广泛的应用前景。首先针对应用场景,统计并归纳了任务载荷双光图像数据集,为相关研究及应用提供数据支持。其次,跟踪深度学习领域的最新技术,系统总结基于深度学习的双光载荷图像融合方法,将现有方法归纳为生成式和判别式两类,并详细梳理代表性的方法及特点。再次,从定性分析、融合质量及运行效率3个方面,对不同类型的图像融合方法在不同数据集上的融合效果进行深入的实验对比分析。最后,探讨图像融合技术在低空遥感应用中面临的挑战问题,为相关领域的研究提供参考。

关键词: 低空遥感, 深度学习, 任务载荷, 双光图像, 图像融合

Abstract:

Dual-band payload image fusion has broad prospects for application in the field of low-altitude remote sensing such as target monitoring, disaster warning, and professional inspection. Firstly, dual-band images datasets of task payload are summarized to provide data support for relevant research and applications. Secondly, by tracking the latest technologies in the field of deep learning, a systematic review of deep learning-based dual-band payload image fusion methods is conducted. These methods are categorized into generative and discriminative approaches, and representative algorithms along with their characteristics are detailed. Thirdly, in-depth experimental comparative analysis of different types of image fusion methods is performed on various datasets, evaluating their performance from three aspects: qualitative analysis, fusion quality, and operational efficiency. Finally, the challenges faced in application of the image fusion technology in low-altitude remote sensing are discussed, offering valuable insights for research in related fields.

Key words: low-altitude remote sensing, deep learning, task payload, dual-band image, image fusion

中图分类号: