干涉成像光谱仪空谱信息同步复原方法  

Simultaneous Spatial and Spectral Information Recovery for Interferometric Imaging Spectrometer

作　　者：朱锋 安军社 施海亮[3] 熊伟[3] 王先华 Zhu Feng;An Junshe;Shi Hailiang;Xiong Wei;Wang Xianhua(Key Laboratory of Electronics and Information Technology for Spacing Systems,National Space Science Center,Chinese Academy of Sciences,Beijing 100190,China;University of Chinese Academy of Sciences,Beijing 100049,China;Anhui Institute of Optics and Fine Mechanics,Hefei Institute of Physical Science,Chinese Academy of Sciences,Hefei 230031,Anhui,China)

机构地区：[1]中国科学院国家空间科学中心复杂航天系统电子信息技术重点实验室,北京100190 [2]中国科学院大学,北京100049 [3]中国科学院合肥物质科学研究院安徽光学精密机械研究所,安徽合肥230031

出　　处：《光学学报》2022年第24期222-231,共10页Acta Optica Sinica

基　　金：国家重点研发计划(2016YFB0500704);中科院重点部署项目(ZDRW-KT-2020-3)。

摘　　要：以干涉高光谱成像模型为基础,提出一种结合先验约束的空谱信息同步复原模型,通过非负低秩特性和全变分(TV)正则项分别约束复原高光谱图像的谱间强相关性和空间分段平滑特性,并采用L1范数和Frobenius范数分别对干涉数据中的稀疏性噪声和高斯噪声进行建模。模拟和真实干涉数据的对比实验验证了所提方法的有效性。相比于传统的干涉数据复原方法,所提方法在准确复原目标光谱信息的同时,能够有效地消除干涉图混合噪声的退化影响,从而提高复原高光谱图像的数据质量。On the basis of the imaging model of an interferometric imaging spectrometer,a novel model with a priori constraints is proposed to recover spatial and spectral information simultaneously.The nonnegative low rank property and total variation(TV)regularization are adopted to constrain the strong spectral correlation and spatial piecewise smoothness of recovered hyperspectral image,respectively.Meanwhile,the sparse noise and Gaussian noise in interferometric data are modeled by the L1 norm and Frobenius norm,respectively.The effectiveness of the proposed method is verified through comparative experiments on simulated and real interferometric data.Compared with traditional recovery methods for interferometric data,the proposed method not only recovers the spectral information of the object accurately but also effectively eliminates the degradation effect of mixed noise in the interferogram.As a result,the data quality of the recovered hyperspectral image is improved.

关 键 词：光谱学 干涉成像光谱仪 光谱复原 低秩 全变分 正则化 

分 类 号：TH744[机械工程—光学工程]