基于压缩感知的高分辨率红外成像理论和方法研究

项目名称： 基于压缩感知的高分辨率红外成像理论和方法研究

项目编号： No.61271440

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 无线电电子学、电信技术

项目作者： 刘昆

作者单位： 中国人民解放军国防科学技术大学

项目金额： 80万元

中文摘要： 高分辨率遥感成像要求探测器具备更小的像素尺寸和更多的像素数，同时产生的高数据率给数据存储和传输系统带来巨大的压力。压缩感知提出一种新的采样理论，以远低于奈奎斯特采样速率采集信号的非自适应线性投影(测量值)，然后通过求解优化问题，精确重构原始信号。基于压缩感知的遥感成像方法研究可为高分辨遥感成像提供一种新的技术途径，从本质上提升现有遥感成像系统性能。 本课题基于压缩感知理论，采用焦平面编码的压缩采样方式，研究高分辨率红外成像方法及其关键技术。重点研究解决以下几个问题：1、基于冗余字典的遥感图像稀疏表示方法；2、满足可重构条件的确定性测量矩阵优化设计；3、焦平面编码的高分辨率红外成像方法；4、红外压缩成像实验验证。通过上述研究，丰富和发展压缩感知理论，提出一种新型的红外压缩成像方法，为压缩感知在遥感压缩成像上的应用打下坚实的理论与技术基础。

中文关键词： 压缩感知；遥感视频成像；稀疏表示；稀疏重构；压缩测量

英文摘要： High resolution remote sensing imaging yields image sensors with smaller pixel-pitch and more pixels. The resulting high data rate brings in considerable burden in terms of data storage and transmission. Compressed Sensing (CS) is a new sampling theory, which captures the non-adaptive linear projections of compressible signals at a rate significantly below the Nyquist rate. These signals are then reconstructed from these projections using an optimization process. Researching the method of remote sensing imaging based on CS provides a new technical approach for high resolution remote sensing imaging, which essentially enhance the performance of remote sensing imaging system. In this project, we study the method and key technologies of high resolution infrared imaging by implementing focal plane coding approach based on CS. The main research works are included as follows: 1, Sparse representation method of remote sensing images based on over-completed dictionary. 2, Optimization design of deterministic measurement matrices which satisfy the reconstructed condition.3, Method of high resolution infrared imaging based on focal plane coding.4, the experimental validation of infrared compressive imaging. Through the above designs, the purpose of the project is that enriching and developing the CS theory and proposing a

英文关键词： compressed sensing；remote sensing video imaging；sparse representation；sparse reconstruction；compressed sampling