光谱成像超分辨率光学编码原理与实现方法研究

High resolution spectral imaging is an urgent requirement to achieve reliable target detection and object identification. Optical encoding superresolution method is proposed to improve resolution of spectral imaging in this research, which has not been proposed before. Compared to the conventional superresolution methods using image processing, this method achieves resolution improvement with optical encoding method, thus avoiding time-consuming computer processing. This research provides a brand-new and effective superresolution approach for spectral imaging. We can expect that this research will have important application in spectral imaging and pushes its development.In this research, principle analysis, model simulation and verification experiment are performed to fully understand the optical encoding method for spectral imaging. Firstly, the resolution of spectral imaging system and its space-bandwidth product are analyzed in multi-degree-of-freedom domain using optical information theory. Space-bandwidth product adaption is analyzed for spectral imaging which acquiring spatial and spectra information simultaneously. Optical encoding in different dimensions to change the space-bandwidth product is studied respectively. Secondly, the simulation is performed to verifying the results of principle analysis. Experiment device is also assembled and verification experiment is used to check the performance of optical encoding superresolution methods for spectral imaging.

本项目基于光学信息理论，通过对多维自由度空间光谱成像模型和光学编码原理分析，研究光学编码光谱成像超分辨率原理,研究基于光谱成像系统信息维度的光学编码方法，克服现有图像处理光谱成像超分辨技术的缺点，实现高分辨率光谱成像。运用原理研究、建模仿真和实验验证等方法从理论与实验两方面入手，建立光谱成像系统超分辨光学编码的物理模型，研究光学编码实现光谱成像系统空间带宽积变换，探索光谱成像系统超分辨光学编码的实现方法，建立相应的原理性实验装置。该项研究是光谱成像技术发展的迫切需求，是一种基于新原理和新方法的光谱成像超分辨率技术，在发展高分辨率光谱成像技术方面具有重要的应用前景，是实现高性能光谱成像的有效途径之一，将有力推动高性能光谱成像技术的发展。

光谱成像技术是上世纪90年代应航天遥感领域的需求而发展起来的一种重要光学成像技术，在对地遥感、空间目标探测识别、伪装识别和化学品检测等领域具有广泛应用。本项目针对光谱成像系统的超分辨问题，研究利用光学编码方法提高光谱成像系统的分辨率。.(1).开展空间编码压缩光谱成像系统数学模型的研究，分析成像系统中光学系统、焦平面探测器等组件对空间和光谱分辨率的影响。提出采用具有亚像元结构特征的光学编码板对目标进行编码，分析探测器像元对编码板的降采样模型，利用迭代算法对获得的图像数据进行运算处理，获得高空间分辨率的光谱图像。改进空间编码压缩光谱成像复原算法，提出在算法中增加表征数据光谱维连续性的变差项，提高光谱分辨率并改善数据重建质量。 .(2).研究焦平面探测器的空间编码技术，克服焦平面探测器像元尺寸对光谱成像系统分辨率的影响，并提高图像重建的运算速度。提出在焦平面探测器上采用灰度编码结构，通过多帧图像重建，复原得到高分辨率图像。利用简单的光学编码结构和传统微扫描方法相结合，提高成像系统的几何分辨率。.(3).针对前置光学系统对空间编码压缩光谱成像系统的影响，研究光学系统的波前复原方法。针对现有波前复原算法运算速度缓慢的缺点，通过推导点扩散函数的解析表达式，提出快速波前复原算法。通过波前复原结果重建光学系统的点扩散函数，并对光谱图像进行维纳滤波，提高光谱图像的分辨率。.(4).建立光学编码超分辨光谱成像实验装置，并分别利用光刻技术和空间光调制器实现光学编码结构，开展相关实验。发表学术期刊/会议论文7篇，其中EI收录6篇，ISTP收录2篇，申请专利3项，累计3人次参加国内/国际学术会议交流。