摘要全景图像是指一种宽视角的图像。随着现在数码技术的成熟，使用一些专门的摄像设备，我们就可以直接拍摄出全景图像。但是这些专门的摄像设备往往因为其价格比较昂贵，很少在日常生活中使用。在目前，大多采用普通相机拍摄多张具有重叠区域的图像，最后通过一定的图像拼接技术，把这些图像拼接成一张全景图像。随着图像拼接技术的发展，拼接后的全景图效果越来越理想。同时，图像拼接的应用范围是非常广泛的，它可以运用在虚拟现实，卫星照片的处理以及医学图像的处理等。85058

本文主要介绍介绍现在存在的比较经典的基于特征点的图像拼接技术和总结图像拼接技术，并且在实际项目中完成了应用，实现了三张数字图像之间的全景图像拼接。本文主要包括以下几个方面：针对传统的Harris角点检测算法，采取了更加高效的方法进行检测，即通过压缩非最大角点响应值，实现较为快速的挑选出Harris角点响应值较大的角点。之后没有采用传统的Harris角点检测之后采用的NCC相似性度量，而是采用了欧式距离平方的方法对Harris角点进行粗匹配。之后同过RANSAC算法对伪配点对进行剔除，并据此计算出变换矩阵，完成全景图像的拼接。

毕业论文关键字： 全景图像；图像拼接；欧式距离；特征提取

Abstract The panoramic image refers to a wide angle image。 With the maturity of the digital technology and the use of some special camera equipment, we can directly take a panoramic image。 But these special camera equipment is rarely used in our daily life because of its relatively high price。 At present, most of the images are taken by ordinary cameras, and then the images are handled into a panoramic image by image mosaic technology。 With the development of image mosaic technology, the quality of the panoramic image is becoming more and more ideal。 At the same time, the application of image stitching is very wide, and it can be used in the virtual reality, the processing of satellite photos and the processing of medical image。

This paper mainly introduces the existing classical image stitching technique based on feature points, and summarizes the image mosaic technology。 Apart from this, we apply this technology into actual project and complete the panoramic image mosaic between three digital images。 This article mainly includes the following aspects: compared to traditional Harris corner detection, more effective method of detection is adopted, through compressing non-maximum angle point response value, which achieves good result。 After the detection of Harris corner, we adopt Euclidean distance square to match the Harris corner instead of the NCC similarity measure。 After that, the RANSAC algorithm is used to eliminate the false matching points, to calculate the transformation matrix, and to complete the image mosaic。 

Key words: panoramic image, image mosaic, Euclidean distance, feature extraction 

目  录

第一章 绪论 1

1。1引言 1

1。2研究课题的背景和意义 1

1。3国内外研究现状 1

1。3。1 国外研究现状 2

1。3。2 国内研究现状 4

1。4论文的主要工作 4

1。5论文的组织结构 4

1。6本章小结 5

第二章 图像拼接技术综述 6

2。1引言 6

2。2 图像处理流程