节点文献
基于点和边缘相结合特征提取的图像配准算法
Image registration algorithm based on extraction of feature point and edge
【摘要】 提出了一种新的基于几何特征的图像配准算法,将特征点提取和边缘提取的方法相结合,有效的获得了图像的主要信息。通过立体视觉匹配技术,建立图像之间的点映射关系,并采用二叉树搜索策略改进搜索效率。根据空间的联立方程组得到图像在两个坐标系统中的三维空间数据坐标,然后进行数据拟合得到非特征点的数据。该方法提高了图像配准的效率和准确性。
【Abstract】 A new algorithm about image registration is presented,which has combined feature point extraction and edge extraction,and has acquired main information of images.A relationship of point-to-point mapping is established by means of stereo vision and a im-proved binary-tree search tactics,so that the accurate data in 3-D space based on space several equations is gotten,then the data of other point is calculated.The efficiency and accuracy are improved obviously.
【关键词】 图像配准;
特征点提取;
边缘提取;
控制点;
图像相关;
【Key words】 image registration; feature point extraction; edge extraction; control point; image correlation;
【Key words】 image registration; feature point extraction; edge extraction; control point; image correlation;
- 【文献出处】 计算机工程与设计 ,Computer Engineering and Design , 编辑部邮箱 ,2009年04期
- 【分类号】TP391.41
- 【被引频次】19
- 【下载频次】815