铁路监控视频的超分辨率技术研究

摘要：超分辨率技术是一种由同一场景下多帧由相互运动的低分辨率图像重建出高分辨率图像的技术。该技术的特点就是可以经由软件算法充分利用低分辨率图像间的信息互补来突破硬件分辨率规格的局限，实现对高分辨率图像的获取。该特点能够有效地应用在铁路监控视频中，解决诸如存储、硬件限制及传输通道不足等诸多问题。本文首先叙述了图像超分辨率重建技术的国内外研究情况，详细描述了超分辨率技术的原理，分析数字图像生成的观察模型，总结了重建过程中的亚像素配准、图像融合和图像恢复三个环节。其次，对在MATLAB实现的几种重建算法的结果优劣之处进行了比较和分析。最后，在本文的结尾部分总结了对超分辨率重建技术的全部学习研究工作，以及对该技术未来研究方向的发展展望。27718
毕业论文关键字：超分辨率技术；观察模型；亚像素配准；图像融合
Super-Resolution Technology In Surveillance Video of Railway
Abstract: The principal of super-resolution technology is that it superposes low resolution images in one scene with mutual movement to achieve a high resolution image in the scene. The feature of this technology is that the information from low-resolution images can be complementarily add up to achieve a high-resolution image, which breaks the limitation of the specification of the resolution from a hardware.With its feature, this technology can be effectively utilized in railway surveillance video, which solves problems such as storage, limitation of hardware and information transmission.This paper first described the research situation of super-resolution image re-construction technology in China and abroad countries, fully illustrated the principal of super-resolution technology, analyzed the observation model generated by digital image, summarized sub pixel registration phase, image fusion phase and image restoration phase in re-construction process.Secondly, the results of several different ways of reconstruction algorithm in MATLAB are compared and analyzed.Finally, the end of this paper summarized the entire study and research work on the super-resolution re-construction technology, as well as the direction of future research and technological development prospects.
KeyWords: Super-resolution technology; Observation model; Sub-pixel registration; Image Fusion
目录
1 绪论    1
1.1 图像分辨率的概念    1
1.2 图像超分辨率重建技术的研究背景及意义    1
    1.2.1研究背景    1
    1.2.2 研究意义    2
1.3 超分辨率技术在铁路监控中的应用    3
1.4 超分辨率技术的国内外研究现状    4
    1.4.1 国外研究概况    4
    1.4.2 国内研究概况    5
1.5 本文的主要工作与章节安排    5
2 超分辨率技术的原理    7
2.1 观测模型的概念    7
2.2 超分辨率重建的重建基础    8
    2.2.1 超分辨重建的基础—亚像素的采样    8
    2.2.2 采样过程中模糊的产生    10
2.3 亚像素配准    12
    2.3.1亚像素配准的概念    12
    2.3.2 亚像素配准简介    12
2.4 章节小结    13
3 常见超分辨率算法介绍    14
3.1 空域法算法    14
    3.1.1 非均匀插值法    14
    3.1.2 迭代反向投影法（IBP）    15 铁路监控视频的超分辨率技术研究:http://www.751com.cn/tongxin/lunwen_22291.html