基于框小波变换的图像融合的算法研究 毕业设计论文完结.doc

本科毕业设计 题目： 基于框架小波变换的图像融合的算法研究 学 院: 信息科学与工程学院 专 业: 电子信息工程专业 学 号: 学生姓名: 指导教师: 日 期: 摘 要 图像融合是多传感器数据融合的一个重要分支。其目的是通过对多幅源图像信息的提取与综合，获得对同一场景或目标的更为准确，更为全面、更为可靠的图像描述。目前图像融合技术已经广泛应用于遥感、医学、计算机视觉、军事等方面。 图像融合从抽象层次上分为：像素级、特征级和决策级图像融合。本论文主要研究像素级图像融合，研究重点是基于框架小波变换的图像融合。首先介绍了图像融合的提出、国内外发展现状以及其广泛的应用；而后介绍了小波变换的理论，多分辨率分析，基于小波系分析的图像融合，继而实现了框架小波变换的图像融合并在此基础上通过实验结果讨论了最佳分解小波基与最佳融合规则；提出了框架小波的局限性并通过基于区域的图像融合方法加以改进。最终探讨了二者相结合的图形融合的新方法，得到了良好的效果。 关键字: 图像融合； 小波变换； 框架小波； 区域变换 Abstract One of the very important branch which belong to Multi-sensor data fusion is Image Fusion.The accurate, safe and reliable demonstration of an image is required, which can be obtained through the extraction and combination of som Source image information.In this period, Image Fusionis already be used broadly in Rocker, medicine, computer vision,military,etc. Image Fusion can be classify from the abstraction level to Pixel level, feature level and decision-making level.Mostly, this thesis is studying the Pixel Level Image Fusion, and the vital point is based on Framework of the wavelet transform.Initially, the propose of Image Fusion, the present situation at home and externally are analyzed, and also the widely usage of it. After that, this thesis shows the theory of Wavelet Transform, multi-resolution analyzing and Image Fusion which based on Wavelets analysis. Then, Framework of wavelet transform fusion is obtained. Consequently, the best way to resolve Wavelet and the best rule to mix them are discussed. Furthermore, this thesis suggests the disadvantage of Frame wavelet and the method which can brings improvement based on Region fusion. Finally, some favorable effects are achieved by the combination of Both which is a new method of Image Fusion. Key words：Image fusion; wavelet transform; Frame wavelet; regional transformation  目 录 1 绪论 1 1.1 图像融合 1 1.2 图像融合的研究意义 2 1.3 图像融合的研究现状 3 1.4 本文的主要内容 3 2 小波变换理论 5 2.1 小波变换的发展 5 2.2 连续小波 5 2.3 小波函数 6 2.4 离散小波变换 8