| 基于二维特征和Fisher准则函数车牌图像分割 |
| |
| 引用本文： | 白莹,谢国庆,王智文.基于二维特征和Fisher准则函数车牌图像分割[J].广西工学院学报,2005,16(4):36-39. |
| |
| 作者姓名： | 白莹 谢国庆 王智文 |
| |
| 作者单位： | 福建工程学院数理系,福建,福州,350007;福建师范大学软件学院,福建,福州,350007;广西工学院计算机工程系,广西,柳州,545006 |
| |
| 摘 要： | 以车牌图像的灰度和邻域均值构成二维特征，此二维特征的直方图具有如下的特点：车牌图像中的字符目标区域与背景区域集中在二维特征直方图中的对角线附近。可方便的运用Fisher准则函数在二维特征直方图的对角线找到最佳分割点。实验表明，相对于直接在二维灰度直方图上进行向量分割的方法，不仅能降低误判率，还可将二维问题转化为一维问题，可大大减少可计算量。
|
| 关 键 词： | 二维特征直方图 Fisher函数 图像分割 |
| 文章编号： | 1004-6410（2005）04-0036-04 |
| 收稿时间： | 2005-09-02 |
| 修稿时间： | 2005年9月2日 |
Car plate image segmentation based on 2_D characteristic and Fisher function |
| |
| BAI Ying,XIE Guo-qing,WANG Zhi-wen.Car plate image segmentation based on 2_D characteristic and Fisher function[J].Journal of Guangxi University of Technology,2005,16(4):36-39. |
| |
| Authors: | BAI Ying XIE Guo-qing WANG Zhi-wen |
| |
| Affiliation: | 1. Department of Mathematics and Physics,Fujian Engineering Unversity, Fuzhou 350007,China; 2. School of Software, Fujian Normal University, Fuzhou 350007 ,China; 3. Department of Computer Engineering ,Guangxi University of Technology, Liuzhou 545006, China |
| |
| Abstract: | The 2__D characteristic of car plate image is composed of the grey of each pixel and the average of neighborhood field. There is the following feature on the histogram of 2__D characteristic: the character object region and background region of image collect in the vicinity of diagonal on the histogram of 2__D characteristic. It is convenient to use Fisher function to search the optimal point on the diagonal of the histogram. Experimental results show that as to the current methods of separating 2_D histogram with a threshold, the new method that uses the Fisher discriminant function to segment image can not only reduce the error rate caused by noise, but also decrease the computational time greatly by converting the 2_D problem to the 1_D problem. |
| |
| Keywords: | 2_D characteristic histogram Fisher function image segmentation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
