摘要在较大目标场景摄影采集时，人们为了冲破角度的局限性，钻研开发出了图像拼接技术。作为数字图像处理研究的一个大势倾向，目前它在全景图的拼接、视频检索压缩、医学图像处理、虚拟现实技术、超分辨率重构等诸多领域的应用实践，都彰显出其广泛研究前景和发展潜力。图像拼接要实现的目的就是是将数张两两之间具有一定重叠部分的图像，拼合组成一幅没有明显缝合线的高分辨率超宽视角图像。25422
本课题首先对图像拼接技术的研究目的和背景、应用领域和国内外研究现状作了简要介绍。主要围绕图像拼接的基本流程和关键技术，对图像预处理、图像匹配算法、图像拼接算法以及图像拼合处理方法进行了叙述。最后对基于SIFT和Harris角点匹配的可见光图像拼接进行了MATLAB语言环境下的仿真，并且对基于特征点、基于傅里叶梅林变换的算法做对比分析。
关键词：图像拼接，Harris算子，SIFT算子，傅里叶梅林变换，图像配准
毕业论文设计说明书外文摘要
Title Design and implementation of image mosaic algorithm
Abstract
In order to break the limitations of the angle in photography collection for large target scene, people developed image mosaic technology. As an important research direction of digital image processing technology. The applications in the mosaic panorama, video retrieval compression, medical image processing, application of virtual reality technology practice, super-resolution reconstruction and many other fields are all showing its extensive research prospects and development potential. The main task is to use more than one picture with certain overlap part to form a large, high resolution and seamless complete picture.
This paper firstly introduced the purpose, the background, the applications and the current situations of image mosaic technology. The main study is about the basic processes and key technologies of image mosaic, such as image matching algorithm, image pre-processing, image stitching algorithm and image processing method. Finally, I do some simulations of the visible image stitching algorithm based on SIFT algorithm and Harris corner matching under the environment of MATLAB language, and have a comparison and analysis of image stitching algorithm based on feature points and image stitching algorithm based on Fourier-Mellin transform to determine the suitable image stitching algorithm for visible images.
Keywords: image mosaic, Harris algorithm, SIFT algorithm, Fourier-Mellin transform, image registration
目录
1 绪论 1
1.1 图像拼接技术研究目的及背景 1
1.2 图像拼接技术的国内外研究现状 2
1.2.1 国外研究现状 2
1.2.2 国内研究现状 2
1.3 图像拼接的的基本流程 3
1.4 图像拼接的算法分类 3
1.5 本文主要的研究工作内容 4
2 图像采集和预处理 5
2.1 图像采集 5
2.2 图像预处理 5
3 图像配准 7
3.1 图像变换模型 7
3.2 图像配准步骤 9
3.3 特征提取 9
3.3.1 基于Harris角点的特征提取 9
3.3.2 基于SIFT（尺度不变特征转换）的特征提取 10
3.3.3 基于傅里叶梅林变换的特征提取 15
3.4 RANSAC算法 16
4 图像的合成与拼接 18
4.1 图像的几何变换 18
4.2 图像插值 18 MATLAB图像拼接算法的设计与实现:http://www.youerw.com/tongxin/lunwen_19210.html