| 基于小波包变换的多尺度图像匹配算法 |
| |
| 引用本文： | 吴纯,王文波.基于小波包变换的多尺度图像匹配算法[J].科技情报开发与经济,2006,16(8):149-150. |
| |
| 作者姓名： | 吴纯 王文波 |
| |
| 作者单位： | 1. 武汉商业服务学院,湖北,武汉,430056
2. 武汉大学,湖北,武汉,430072 |
| |
| 摘 要： | 针对梯度算子的图像匹配方法存在收敛速度慢、迭代容易落入局部最优点等问题,提出了利用小波包的多尺度特征对图像进行不同分辨率的分解的方法。实验表明基于小波包变换的多尺度图像匹配算法准确度高,计算速度较快且减小了迭代落入局部最优解的概率。
|
| 关 键 词： | 小波包变换 多尺度图像匹配 收敛速度 |
| 文章编号： | 1005-6033（2006）08-0149-02 |
| 收稿时间： | 2005-11-28 |
| 修稿时间： | 2005年11月28 |
The Multiple-dimensional Image Matching Algorithm Based on the Wavelet Packet Transformation |
| |
| WU Chun,WANG Wen-bo.The Multiple-dimensional Image Matching Algorithm Based on the Wavelet Packet Transformation[J].Sci-Tech Information Development & Economy,2006,16(8):149-150. |
| |
| Authors: | WU Chun WANG Wen-bo |
| |
| Abstract: | In the light of the problems of the slow rare of convergence and that the iteration falls easily onto the locally optimal points existing in the image matching method of the gradient operator,this paper puts forward a method for making the decomposition with different resolutions for the image by using the multiple-dimensional characteristic of the wavelet packet.The test shows that the multiple-dimensional image matching algorithm based on the wavelet packet transformation,which possesses higher accuracy and higher speed,can reduce the probability of that the iteration falls onto the locally optimal points. |
| |
| Keywords: | wavelet packet transformation multiple-dimensional image matching rare of convergence |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
