|
|||||
|
|
| 一种混杂的图像配准方法 | |
| 引用本文： | 柴玉华,王蓉,高立群,田磊,王坤.一种混杂的图像配准方法[J].控制与决策,2006,21(7):755-758. |
| 作者姓名： | 柴玉华 王蓉 高立群 田磊 王坤 |
| 作者单位： | 1. 东北大学,信息科学与工程学院,沈阳,110004;东北农业大学,工程学院,哈尔滨,150030 2. 东北大学,信息科学与工程学院,沈阳,110004 3. 中芯国际北京集成电路制造有限公司,北京,100176 |
| 基金项目： | 公安部重点科研项目（20029322301）;黑龙江省自然科学基金项目（F0318）. |
| 摘 要： | 提出一种基于特征和亮度的混杂图像配准方法．首先对源图像进行小波分解，在小波金字塔最顶层的低频近似子图像上进行角点检测，通过仿射变换模型得到初始配准参数，然后将这一参数逐层进行迭代精炼，直至达到金字塔最底层，得到最终精确的配准参数．将这些不同配准方法结合起来，可以克服单独使用任一方法存在的不足,理论分析和实验结果表明，这种混杂图像配准算法精度较高． |
| 关 键 词： | 图像融合 图像配准 角点检测 多尺度 |
| 文章编号： | 1001-0920（2006）07-0755-04 |
| 收稿时间： | 2005-08-01 |
| 修稿时间： | 2005-08-012006-02-28 |
A Hybrid Image Registration Algorithm |
|
| CHAI Yu-hua,WANG Rong,GAO Li-qun,TIAN Lei,WANG Kun.A Hybrid Image Registration Algorithm[J].Control and Decision,2006,21(7):755-758. | |
| Authors: | CHAI Yu-hua WANG Rong GAO Li-qun TIAN Lei WANG Kun |
| Affiliation: | 1. College of Information Science and Engineering, Northeastern University, Shenyang 110004, China; 2. College of Engineering, Northeastern Agricultural University, Harbin 150030, China; 3. Semiconductor Manufacturing International Corporation, Beijing 100176, China |
| Abstract: | A hybrid image registration algorithm based on character and intensity is proposed. Source images with wavelet are decomposed, and corner information can be detected in the two approximate sub-images on the wavelet coarsest level. The initial registration parameters are achieved through affine transform model. The optical flow is iteratively refined to get the coarse registration parameters. The process will continue until the final and accurate matching parameters are obtained. The combination of these different methods tends to compensate for any deficiencies in the individual methods. Theory analysis and experimental results show that the performance of this method is good. |
| Keywords: | Image fusion Image registration Corner detection Multi-scale |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |
| 点击此处可从《控制与决策》浏览原始摘要信息 | |
| 点击此处可从《控制与决策》下载全文 | |