高空间分辨率电子倍增超光谱成像信噪比评价与测试方法研究

Hyperspectral imaging has characteristic of fast frame rate, low signal noise ratio at high spatial resolution. At present, the oscillating mirror was usually used to reduce the application requirements, but it increases the volume and weight, image obtained is not continuous, and the moving parts reduce the reliability of spaceflight. The project combined high-speed electron multiplication with imaging spectrum to solve the contradiction between less light energy and high transfer rate under high resolution. This study proposed that taking use of new imaging control method to construct high spectrum imaging system based on EMCCD. The modulation law of multiplication gain and adaptive correction mechanism would be investigated, signal- to- noise model of spectra related to energy was established according to the spectral information transfer and transform, the overall signal-to-noise ratio optimization method of different spectrum, different seasons and different latitude was discussed. At last, the evaluation method of signal-to-noise ratio was studied. The theoretical knowledge, experimental data and technical experience were accumulated for researching on practical system of high spatial and spectral resolution electronic multiplying hyperspectral imaging system in future.

超光谱成像在高空间分辨率下帧速率高、信噪比低，现今多采用摆镜降低应用要求，增加了体积和重量，获取图像不连续，且运动部件降低了航天可靠性。本项目将超高速电子倍增与成像光谱有机结合，解决高分辨率下低光能量与高转移速率的矛盾。本研究提出利用新成像控制方法，拟构建基于电子倍增电荷耦合器件（EMCCD，Electronic Multiplying CCD）的高分辨率超光谱成像实验系统；探讨倍增增益的调制规律与自适应校正机制，从光谱信息传递和变换全环节建立与光谱和能量相关的信噪比模型，探讨不同谱段不同季节及不同纬度下的整体信噪比优化方法，最后进行信噪比评价方法的研究，为将来研制实用的高空间分辨率电子倍增超光谱成像系统积累理论知识、实验数据和技术经验。

EMCCD以其体积小、灵敏度高的特点，在高空间和光谱分辨率超光谱对地观测成像中展示出良好的应用前景。本项目突破了光谱和空间分辨率的能量约束，建立受太阳高角、地物反射率、光学系统参数、光谱宽度、地面像元分辨率等约束的信噪比模型，转变当前“摆镜构造相对静止”的信噪比提高思想，建立了基于电子倍增信噪比提升理论，并引入基于EMVA1288的信噪比评价方法；通过不同谱段的光谱信号选取不同的倍增增益，并对每个光谱信号根据地物反射率对应的电子数判断是否电子倍增，实现不同谱段光谱图像的整体动态范围提升；搭建了实验室测试平台，通过实验测试了焦面辐照度在0.465mlx～0.2lx范围内探测器的电子倍增模式下的信噪比，通过选择适当的电子倍增增益，信噪比可提高6倍，验证了弱光谱辐照度下无摆镜的信噪比提高可行性，并利用打印在单张纸片的不同颜色的文字进行了光谱图像电子倍增试验，通过调节电子倍增增益获取了信噪比良好的光谱图像。