| 基于复小波变换的图像镶嵌 |
| |
| 引用本文： | 宋利,周源华,周军.基于复小波变换的图像镶嵌[J].上海交通大学学报,2005(Z1). |
| |
| 作者姓名： | 宋利 周源华 周军 |
| |
| 作者单位： | [1]上海交通大学图像通信与信息处理研究所 [2]上海 |
| |
| 基金项目： | 国家自然科学基金资助项目(69905003) |
| |
| 摘 要： | 提出了一种新的图像镶嵌算法,首先根据图像采样的几何约束和相关特性,求得图像序列的初始拼接区域.然后采用满足平移不变性的复数小波变换对相邻图像序列进行了多分辨率的匹配,在每一层分辨率下,均采用8参数的投影变换模型进行了图像的变换.该方法将图像的匹配和摄像机的标定统一在一起,采集方法简单,既能保证匹配的精度,又提高了运算的速度.实验结果表明,图像合成效果非常理想.
|
| 关 键 词： | 图像镶嵌 复小波变换 整合变换 |
Image Mosaic Based on Complex Wavelet Transform |
| |
| SONG Li,ZHOU Yuan-hua,ZHOU Jun.Image Mosaic Based on Complex Wavelet Transform[J].Journal of Shanghai Jiaotong University,2005(Z1). |
| |
| Authors: | SONG Li ZHOU Yuan-hua ZHOU Jun |
| |
| Abstract: | A novel panorama composition algorithm was presented. According to geometry constraints and correlation property, the initial stitching area of image sequences is solved. Then shift-invariant complex wavelet transformation is employed to match image sequences in multi-resolution. In every resolution, eight-parameter projective transformation model is used to transform images. The matched images and calibrated camera are united by the algorithm at the same time, which is easily operated. It can not only ensure the matching precision but also improve the operation speed. The result of experiments shows that the effect of image composition is ideal. |
| |
| Keywords: | image mosaic complex wavelet translation collineation transformation |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
