图像融合技术的研究.doc

***大学毕业设计（论文） PAGE PAGE \* MERGEFORMAT III 摘要 图像拼接是将存在相互重叠部分的图像进行空间匹配对准，经重采样后形成一幅新图像的技术，该技术在遥感图像处理、计算机视觉、医学图像分析、军事等领域有着广泛的应用价值。 一般来说，图像拼接主要包括图像配准和图像融合，其中图像配准是整个图像拼接的基础。目前提出的图像配准方法可分为三类，基于灰度信息的图像配准方法、基于变换域的图像配准方法和基于特征的图像配准方法。本文重点研究了基于互信息的图像配准方法，它不需要对图像进行分割等预处理，可以用于任何不同模态的图像配准，特别是在医学图像配准中得到了广泛的应用。图像融合的目的是为了消除拼接缝，目前主要有基于空间域和基于频率域的融合方法。本文主要分析了基于小波变换的图像融合方法，由于小波变换多分辨率图像分析理论的特点，该方法取得了较好的融合效果。最后分别设计图像配准和图像融合的算法，并用Matlab仿真实现，完成图像拼接。 关键词：图像拼接 图像配准 互信息 图像融合 Abstract Image stitching is a technology that carries on the spatial matching to a series of images which overlap with each other, and finally forms a new image after resampling. It has widely applications in the fields of remote sensing image processing, computer vision, medical image analysis, military and so on. Generally speaking, the image stitching process consists of image registration and image fusion, and image registration is the important foundation of image stitching. At present, the method of image registration can be divided into three categories, the grayscale information-based image registration, the transform domain-based image registration and feature-based image registration. This paper focuses on the mutual information-based image registration. It doesn’t require preprocessing for images such as segmentation. It can be used for any modal image registration, and widely applied in the medical image registration. Image fusion is to eliminate the seam between images. It mainly includes spatial domain-based method and frequency domain-based method. This paper mainly analyzes the image fusion method based on wavelet transform, because of the characteristics of the multi-resolution analysis of the wavelet transform, and this method has obtained better effect. Finally the article designs the algorithms of image registration and image fusion, and completes the image stitching by using Matlab. Keywords: Image stitching Image registration Mutual information Image fusion 目录 TOC \o 1-3 \h \z \u HYPERLIN