| 郑鹏,郑可妮,施锡泉.一种基于物体质心的图像自动化校准方法[J].数学研究及应用,2019,39(6):619~632 |
| 一种基于物体质心的图像自动化校准方法 |
| A Centroid-Based Automatic Image Registration Method |
| 投稿时间：2019-08-18 修订日期：2019-10-25 |
| DOI：10.3770/j.issn:2095-2651.2019.06.009 |
| 中文关键词: 自动匹配 阀值函数 封闭边界提取矩阵 图像校准 |
| 英文关键词:automatically matching threshold function closed boundary extracting matrix image registration |
| 基金项目:陕西省教育厅科研计划项目(Grant No.18JK0161),陕西理工大学校级科研项目(Grant No.SLGQD1807). |
|
| 摘要点击次数: 837 |
| 全文下载次数: 532 |
| 中文摘要: |
| 图像校准是将任意相邻两图像都具有公共部分的一列图像合成一副全图的图像处理过程.而图像自动化校准的难点在于对一组图像公共部分特征信息的提取和自动匹配,尤其是自动匹配和图像校准的误差分析.本文基于放射变换不变性,给出了一种基于特征信息质心的自动匹配方法,与目前国际上主流特征提取和特征校准方法相比,本文方法较准确,且适用于卫星图片及一些MRI和CT 图片.本文方法的另一优点是具有线性计算复杂度. |
| 英文摘要: |
| Image registration is a kind of image processing that merges a series of pictures, any two neighbors having a common part, into one global image. Together with the feature extracting, the most difficult work in automatic registration is how to efficiently match the common feature information in different images. Based on the invariant property of affine transform, a centroid-based method of automatic feature matching is presented in this paper. Compared with the current state-of-the-art methods, the method in this paper is more efficient for the testing images, including MRI, CT, and satellite images. Another advantage of the method in this paper is that it has a linear computational complexity. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
