光学遥感图像的超分辨率处理技术综述  

A Survey on the Optical Remote Sensing Image Super-resolution Technology

作　　者：张震洲 高昆[1] 李维[2] 王俊伟 陈卓一 吴穹 苏云 ZHANG Zhenzhou;GAO Kun;LI Wei;WANG Junwei;CHEN Zhuoyi;WU Qiong;SU Yun(Key Laboratory of Photoelectronic Imaging Technology and System,Ministry of Education,Beijing Institute of Technology,Beijing 100081,China;Beijing Institute of Space Mechanics&Electricity,Beijing 100094,China;Institute of Spacecraft System Engineering,CAST,Beijing 100094,China)

机构地区：[1]北京理工大学光电成像技术与系统教育部重点实验室,北京100081 [2]北京空间机电研究所,北京100094 [3]中国空间技术研究院总体部,北京100094

出　　处：《航天返回与遥感》2020年第6期21-33,共13页Spacecraft Recovery & Remote Sensing

基　　金：国家自然科学基金(61875013);装备预研航天科技联合基金(6141B061004)。

摘　　要：光电成像系统探测器的欠采样效应、光学系统的衍射极限等诸多降晰因素影响着光学对地观测系统的分辨力。超分辨率处理技术利用信号处理的手段重建成像系统截止频率之外的信息,可以在不改变硬件设备的前提下,获得高于成像系统分辨率的图像,对于光学遥感器的设计和应用有着重要意义。文章结合了光学遥感图像退化模型和图像超分辨率处理的技术实施途径,按照基于图像重建的方法,基于学习的方法和重建与学习混合的方法进行了分类综述,对比了各种主要算法的基本特点,并对未来的发展方向进行了展望。The under-sampling effect of detectors in the photoelectric imaging system,the diffraction limit of the optical system,and many other degradation factors affect the resolving power of the optical earth observation system.The super-resolution technology can reconstruct the information beyond the cut-off frequency of the imaging system through signal processing,and can obtain images with higher resolution without changing the hardware,which is of great significance to the design and application of optical remote sensors.Combining the optical remote sensing image degradation model and the implementation method of the super-resolution technology,the basic characteristics of the various super-resolution algorithms are summarized and compared according to the classification method based on image reconstruction,the learning-based method and hybrid methods.Also,a prospect for the future development direction is given.

关 键 词：遥感图像 退化模型 超分辨率处理方法 评价指标 航天遥感 

分 类 号：P407.8[天文地球—大气科学及气象学]