摘要
目的提出一种基于小波分析的车牌定位方法,保证车牌定位的准确性,满足车牌识别系统的实时性要求.方法首先对原始图像进行预处理,去除噪声和增强对比度,然后利用原图像的垂直梯度的水平投影实现车牌图像的粗定位,最后利用小波分析提取车牌图像的高频信息,实现车牌精定位.结果通过先粗定位再细定位的方法提高了车牌定位的速度,并且抗噪声能力强,定位效果较好.结论实验证明了该方法的有效性和实用性,为车牌识别系统的后续处理提供了有利条件.
In ordrder to guarantee localization accuracy and meet real-time ability of license plate recognition system, a method is presented based on wavelet analysis to locate license plates. The original images are processed to eliminate the noise and enhance the images. Then rough localization of license images is carried out by level projection of vertical gradients of original images. High frequency information is extracted from license image using wavelet analysis to realize precision localization. As a result, the speed of license localization is improved by rough localization and precision localization, which also results in strong anti-noise ability and high localization accuracy. The experiment results show that the method presented is effective and practical. It can provide better results for post-processing of license recognition system.
出处
《沈阳建筑大学学报（自然科学版）》
CAS
北大核心
2009年第6期1217-1220,共4页
Journal of Shenyang Jianzhu University：Natural Science
基金
国家自然科学基金项目(60874103)
关键词
定位算法
图像处理
小波分析
车牌
localization algorism
image processing
wavelet analysis
license plate
