折反射全向非均匀压缩成像技术研究

项目名称： 折反射全向非均匀压缩成像技术研究

项目编号： No.61275016

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： 谭树人

作者单位： 中国人民解放军国防科学技术大学

项目金额： 78万元

中文摘要： 本项目从提高折反射全向成像系统成像分辨率出发，研究基于压缩感知的折反射全向压缩成像及超分辨率重构方法。首先，研究设计适用于超分辨率重构的折反射全向压缩成像系统光学结构，建立压缩成像模型，分析压缩成像的性能。其次，提出适合折反射全向压缩成像的非均匀测量矩阵，并对测量矩阵的优化方法进行研究。最后，提出折反射全向非均匀超分辨率重构方法。本项目的特色与创新之处在于将压缩感知原理引入折反射全向成像中，采用非均匀测量矩阵压缩成像，通过超分辨率重构同时解决折反射全向图像相对分辨率低和内外环分辨率不均匀的问题。研究成果将扩展全向视觉理论研究和实际应用空间，对折反射全向成像系统在军事远景侦察、机器人视觉导航、全景视频监控、虚拟场景漫游等众多领域的广泛应用产生积极的促进作用。

中文关键词： 折反射全向成像；压缩成像；编码孔径；超分辨率；

英文摘要： To improve the imaging resolution of catadioptric omnidirectional，this project researches on catadioptric omnidirectional compressive imaging and super-resolution reconstruction based on compressed sensing. Firstly, the optical structure of catadioptric omnidirectional is designed for super-resolution reconstruction, the compressive imaging model is builded for analyzing the performance of compressive imaging. Secondly, a nonuniform measurement matrix is proposed for catadioptric omnidirectional compressive imaging and the optimization method of measure matrix is studied. Finally, the method for nonuniform super-resolution reconstruction is proposed. The characteristics and innovation of this project lie in apllying the principles of compressed sensing for catadioptric omnidirectional imaging, compressive imaging with nonuniform measurement matrix, solving simultaneously the low image resolution and the nonuniform resolution between external ring and internal ring in omnidirectional image through super-resolution reconstruction. Research results will significantly expand the space of omnidirectional vision theoretical research and practical application, and will further promote the application space of catadioptric omnidirectional imaging system in many fields, such as in military reconnaissance, robots naviga

英文关键词： Catadioptric omnidirectional image；compressive imaging；coded aperture；superesolution；