一种车牌字符定位分割算法研究  

Research on a license plate character location and segmentation algorithm

作　　者：钟彩[1] 彭春富[1] 傅波[1] 胡常乐 Zhong Cai;Peng Chunfu;Fu Bo;Hu Changle(Changde Vocational Technical College,Changde 415000,China)

机构地区：[1]常德职业技术学院,湖南常德415000

出　　处：《无线互联科技》2022年第18期132-134,146,共4页Wireless Internet Technology

基　　金：2021年湖南省教育厅科学研究项目,项目名称:复杂背景下车牌识别关键技术研究,项目编号:湘教通[2021]352号,21C0962。

摘　　要：近些年,车牌识别技术在智能交通领域中发挥着重要作用,适用于停车场出入管理、高速公路监管及交通部门违章检测等方面,应用前景一片大好,得到社会公众的高度重视。传统的车牌定位算法对夜间车牌图像定位效果整体较差,文章提出了以支持向量机为基础的定位办法。为有效解决夜间环境中汽车车牌图像灰度分布不够均匀的实际问题,运用局部二值拟合模型(LBF)方法执行字符分割过程。统计实验结果发现,以上这种方法能实现精准定位,且能提供精准的光滑闭合式边界,精准度达亚像素级,系统的辨别精准度处于较高层次,有较高推广价值。In recent years,license plate recognition technology has played an important role in the field of intelligent transportation.It is applicable to parking lot access management,expressway supervision and traffic department violation detection.It has a promising application prospect and has been highly valued by the public.The traditional license plate location algorithm has a poor overall effect on the night license plate image location.This paper proposes a location method based on support vector machine.In order to effectively solve the practical problem that the gray distribution of vehicle license plate image is not uniform at night,the local binary fitting model(LBF) method is used to perform character segmentation.After the statistical experiment results,it is found that the above method can achieve accurate positioning,and can provide accurate smooth closed borders.The accuracy can reach the sub-pixel level.The identification accuracy of the system is at a higher level,which is of high promotion value.

关 键 词：夜间图像 车牌定位 支持向量机 局部二值拟合模型 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] U495[自动化与计算机技术—计算机科学与技术]