| 基于特征点的对数极坐标变换图像配准算法 |
| |
| 引用本文： | 雷凯,刘艳滢,王延杰,邢忠宝,尹立敏.基于特征点的对数极坐标变换图像配准算法[J].光学技术,2007,33(3):435-437. |
| |
| 作者姓名： | 雷凯 刘艳滢 王延杰 邢忠宝 尹立敏 |
| |
| 作者单位： | 1. 中国科学院,长春光学精密机械与物理研究所,长春,130033;中国科学院研究生院,北京,100039
2. 中国科学院,长春光学精密机械与物理研究所,长春,130033 |
| |
| 摘 要： | 介绍了一种新的图像配准算法,可以很好地解决图像配准中的平移、旋转和缩放问题。算法的实现是首先使用SUSAN算子提取两幅图像的特征角点,剔除虚假的特征点(噪声点),然后使用改进对数极坐标变换和投影相关匹配算法实现特征点匹配;结合亚像素定位技术,可以进一步提高算法的精度。
|
| 关 键 词： | 图像配准 SUSAN算法 改进对数极坐标变换 亚像素定位 |
| 文章编号： | 1002-1582(2007)03-0435-03 |
| 收稿时间： | 2006/5/19 |
| 修稿时间： | 2006年5月19日 |
Log-polar transformation based on feature points for image registration |
| |
| LEI Kai,LIU Yan-ying,WANG Yan-jie,XING Zhong-bao,YIN Li-min.Log-polar transformation based on feature points for image registration[J].Optical Technique,2007,33(3):435-437. |
| |
| Authors: | LEI Kai LIU Yan-ying WANG Yan-jie XING Zhong-bao YIN Li-min |
| |
| Abstract: | A new algorithm about image registration was presented,which solved translations,rotations and scale transformations problems in image registration process well.The SUSAN algorithm was used to obtain the feature points(interesting points) of the two images.The bad-feature points(the noise points) were eliminated.The improved log-polar transformation and projection correlation matching algorithm were used to acquire the registration parameter of the two images.The sub-pixel location technique was used to improve the registration precision. |
| |
| Keywords: | image registration SUSAN algorithm improved log-polar transformation sub-pixel location |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
