| 模式识别中的总体和分体维纳滤波 |
| |
| 引用本文： | 郑永凯,张凌,董守斌. 模式识别中的总体和分体维纳滤波[J]. 计算机工程与应用, 2001, 37(22): 35-37,66 |
| |
| 作者姓名： | 郑永凯 张凌 董守斌 |
| |
| 作者单位： | 华南理工大学网络中心, |
| |
| 基金项目： | 广东省自然科学基金项目(编号:980603)，高等学校骨干老师资助计划项目(编号:教技司200065号) |
| |
| 摘 要： | 将统计模式识别和统计信号处理进行比较可以看出,对所用的一些统计工具而言,这两个领域的数据形式是相似的,因此信号处理的某些思想也可用于模式识别。文献1初步推导出总体维纳滤波的形式,该文进一步开展了分体维纳滤波。实验中,从原型模式由滤波后产生的位移了解其实际行为,并描绘了相应的总体和分体滤波特性曲线;实际识别效果符合预期估计,聚集度揭示其直接原因。
|
| 关 键 词： | 图像检索 模式识别 特征空间 维纳滤波 |
| 文章编号： | 1002-8331-(2001)22-0035-03 |
General and Individual Wiener Filtering in Pattern Recognition |
| |
| Zheng Yongkai Zhang Ling Dong Shoubin. General and Individual Wiener Filtering in Pattern Recognition[J]. Computer Engineering and Applications, 2001, 37(22): 35-37,66 |
| |
| Authors: | Zheng Yongkai Zhang Ling Dong Shoubin |
| |
| Abstract: | According to the comparison between statistical pattern recognition and statistical signal processing,it can be seen that the organization forms of data in the two fields are similar for being dealt with similar statistical tools,there-fore,some approaches on signal processing can also be applied to pattern recognition.The so-called general Wiener fil-tering method has been firstly presented in ,and a new method named individual Wiener filtering is presented in this paper.In the experiment here,the displacements of prototype models caused by filtering show the actual performances of filtering,and the related general/individual filtering characteristics are shown respectively.The recognition rate is in ac-cord with the expectation,and the crowding level shows its cause. |
| |
| Keywords: | image retrieval pattern recognition eigenspace Wiener filtering. |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
