静态计算光谱成像仪图谱反演的关键数据处理技术  

Critical data processing technology for spectral image inversion in a static computational spectral imager

作　　者：刘扬阳[1] 吕群波[1] 曾晓茹[1] 黄旻[1] 相里斌[1] 

机构地区：[1]中国科学院光电研究院,计算光学成像技术重点实验室,北京100094

出　　处：《物理学报》2013年第6期17-24,共8页Acta Physica Sinica

基　　金：国家高技术研究发展计划(批准号:2011AA7012022)资助的课题~~

摘　　要：静态计算光谱成像技术中图谱反演环节是实现其理论优势极为关键的一步,是决定最终获得图谱质量优劣的数据处理技术.本文为此专注于计算光谱的图谱反演环节,对图像压缩感知理论算法、图像重构算法、以及针对图谱三维数据的反演算法都开展了深入的研究和比较,并结合所研制系统的图谱数据传输全链路和工程研制过程中误差等因素进行全面详尽的仿真验证,给出各种图谱反演算法验证、分析结果.指出静态计算光谱成像系统研制中图谱反演环节的关键数据处理问题,适合采用的算法及其优化路线.为顺利研制静态计算光谱成像仪,保证其理论优势的实现,提供了详实的分析、参考依据.To carry out spectral image inversion in a static computational spectral imager is a crucial step for accomplishing its theoretical advantages, so the data processing technology for spectral image inversion will determine the final spectral image achieved. Focusing on the spectral image inversion, we have investigated various algorithms such as image reconstruction, image compressed sensing and spectral image inversion theories, and compared them carefully. By taking into account the data transmission link of the system and the error in the engineering development process, a comprehensive simulation is carried out. The key issue of spectral image inversion, and also how to use the inversion algorithms to achieve its optimized routes are pointed out. So a detailed analysis for realizing the theoretical advantages and ensuring instrument technology development is provided.

关 键 词：计算光谱 编码孔径 压缩感知 图谱反演 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]