| 基于局部小波分形特征的车牌定位研究 |
| |
| 引用本文： | 贾志勇,何通能,王其聪,李国军.基于局部小波分形特征的车牌定位研究[J].计算机仿真,2005,22(9):208-210. |
| |
| 作者姓名： | 贾志勇 何通能 王其聪 李国军 |
| |
| 作者单位： | [1]浙江工业大学信息工程学院,浙江 杭州 310032 [2]浙江大学信电系,浙江 杭州 310027 [3]浙江省公安高等专科学校公共基础部,浙江 杭州 310053 |
| |
| 基金项目： | 浙江省自然科学基金项目(M603165) |
| |
| 摘 要： | 提出了一种新的车牌定位算法.车牌定位是车牌识别系统中最为关键的处理之一,分形特征能较好地反映牌照区“粗糙”的特性.首先提取牌照图像灰度图的分形维数,再充分利用小波与分形的密切联系,以二次B-样条小波对车牌局部图像进行小波变换后,提取各个子图的分形维数.这些分形维组成的特征矢量能有效地标示车牌区域和非车牌区域,根据这些局部分形维特征,以改进的BP神经网络作为分类器,有效地实现了车牌的快速准确定位,定位时间在0.7s左右,比同类方法更优.
|
| 关 键 词： | 车牌定位 小波变换 分形 分形维 神经网络 |
| 文章编号： | 1006-9348(2005)09-0208-03 |
| 修稿时间： | 2004年6月28日 |
Research on Location of Vehicle-license-plate Based on Wavelet Transform,Fractal Characters and Improved BP Neural Network |
| |
| JIA Zhi-yong,HE Tong-neng,WANG Qi-cong,LI Guo-jun.Research on Location of Vehicle-license-plate Based on Wavelet Transform,Fractal Characters and Improved BP Neural Network[J].Computer Simulation,2005,22(9):208-210. |
| |
| Authors: | JIA Zhi-yong HE Tong-neng WANG Qi-cong LI Guo-jun |
| |
| Affiliation: | JIA Zhi-yong~1,HE Tong-neng~1,WANG Qi-cong~2,LI Guo-jun~3 |
| |
| Abstract: | The paper proposes a new method of vehicle-license-plate location.Vehicle-license-plate location is one of the most important steps of license plate recognition.The character of fractal dimension reflects the roughness of the license-plate region.Extracting the fractal dimension of the gray scale image at first,then,using the close relation between wavelet and fractal to extract the fractal dimension of each sub vehicle-license-plate image after quadric B-spline wavelet transform.These characters can mark the license-plate-region and non-license-plate-region well.With these characters and BP neural network classifier,we can locate the license plate rapidly.This method surpasses other similar methods. |
| |
| Keywords: | Location of vehicle-license-plate Wavelet transform Fractal Fractal dimension Neural network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
