基于SIFT特征向量的图像拼接技术研究学位论文推荐.docxVIP

基于SIFT特征向量的图像拼接技术研究摘要图像拼接技术是数字图像处理邻域的一个研究热点，一直是计算机视觉、模式识别、医学等领域研究的一个重要课题，图像拼接技术也是图像处理工作中的关键技术之一。所谓图像拼接就是将有重叠的图像无缝拼成一幅大宽视域图像的技术。它包含两个关键技术：图像配准和图像融合。近年来，随着技术的成熟，图像拼接技术被很好的应用到了机器人导航、无人平台战场监控、航拍图像处理等多个领域。基于特征的图像配准与拼接技术配准结果准确拼接效果良好且不易受光照、旋转等因素的影响是当前图像配准与拼接领域研究的热点。本文在深入研究和学习已有的基于SIFT的图像配准与拼接技术的基础上，详尽地分析了现有算法的不足，并提出了若干改进算法。关键字：图像拼接，特征，线段特征，图像融合Image splicing technology research based on SIFTfeature vectorabstractImage splicing technology is a research focus in the neighborhood of digital image processing, has been the computer vision, pattern recognition, an important hot topic in the field of medicine and other fields, image splicing technology is one of the key technologies of image processing work. The so-called image mosaicing is there will be overlapping images seamless Mosaic a big wide horizon image technology. It includes two key techniques: image registration and image fusion. In recent years, with mature technology, image splicing technology is very good application in robot navigation, unmanned platform battlefield monitoring, aerial image processing and other fields. Based on the characteristics of image registration and Mosaic registration results are accurate stitching effect is good and not easily influenced by factors such as illumination, rotation is the hotspot in research of image registration and Mosaic. Based on the in-depth research and study of the existing image registration based on SIFT and splicing technology, on the basis of the shortage of the existing algorithm is analyzed in detail, and puts forward some improved algorithm.The keyword：Image stitching, features, line features, image fusion目录第一章绪论31.1研究背景和意义31.2国内外研究现状41.3研究内容5第二章图像拼接的相关理论基础62.1图像拼接的基本流程62.2图像预处理72.3图像配准72.3.1基于灰度信息的图像配准82.3.2基于变换域的图像配准82.3.3基于特征的图像配准92.4图像融合10第三章基于 SIFT 特征的图像拼接算法和图像特征提取算法113.1算法流程113.1.1SIFT 特征描述符113.1.2ISIFT 算法的特征匹配123.2图像融合133.3图像特征提取算法153.3.1图像特征提取概述163.3.2 SIFT 特征向量的匹配16总结17参考文献19第一章 绪论1.1研究背景和意义图像拼接就是将两幅或多幅输入图像通过计算机技术拼合成一幅光滑无缝图像的过程。图像拼接将数张有重叠部分的图像（可能是不同时间