随机编码感知的高分辨遥感光谱计算成像

High-resolution Computational Spectral Imaging of Remote Sensing Based on Coded Sensing

传统遥感光谱成像系统的空间分辨率和光谱分辨率是衡量成像品质的重要指标,而它们受制于探测传感器的点阵密度,提高遥感光谱成像传感器的性能代价非常巨大。另一方面,在入射光能量一定的条件下,由于高分辨光谱的窄波段成像与低窄带辐射能量接收之间的矛盾,传统遥感光谱成像方法的空间分辨率和光谱分辨率往往难以兼得。如何利用普通探测器获得高分辨率图像是众多学者们面临的挑战。针对此问题,文章以压缩感知为理论依据,提出一种基于编码感知—特征解耦的光谱计算成像新方法。该方法无需提高传感器阵列密度,只需控制电子快门曝光和数据处理,就可以做到在保持较高光谱分辨率的情况下,大幅提升空间分辨率。

Spatial resolution and spectral resolution of the remote sensing are the key performance indexes of the remote spectral images, which are restricted to the density of the spectral detector. However, the cost of manufacturing high-density detector is huge. On the other hand, when the energy of the incident light is fixed, high spatial resolution and high spectral resolution of traditional methods usually cannot be acquired simultane- ously ,due to the contradiction between narrow band imaging and low energy received from the narrow spectral bands. How to simultaneously enhance the spatial resolution of the images from the available detectors is facing challenges. Aiming at this problem, this paper, in the principle of compressive sensing, presents a new method of calculating spectral imaging based on coded sensing and characteristics decoupling. The proposed method can greatly enhance the spatial resolution and keep high spectral resolution simultaneously without increasing density of original imaging detector.