车牌图像分类识别技术研究  

Research on license plate image classification and recognition technology

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》2023年第7期97-100,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:21C0962。

摘　　要：在现阶段的交通管理领域,普遍应用车牌识别系统是交通信息化的一个重要组成部分。为了提高车牌图像识别技术应用的效果,文章围绕车牌图像分类识别技术做出分析,在保证车牌图像识别精度基础上提高识别的效率。本文首先介绍车牌图像分类识别技术,了解该技术基本情况;其次介绍车牌图像采集技术、车牌图像特征值提取与分类器、车牌图像处理技术3种车牌图像分类识别的常见技术,了解不同技术在车牌图像分类识别中的应用要点;最后提出加大采集图像内容与质量控制力度、建立车牌识别样本数据库、明确车牌图像识别规范3点建议,明确今后车牌图像分类识别技术的发展方向,以期能够为今后车牌图像分类识别的发展夯实基础。Vehicle license plate recognition system is widely used in the field of traffic management at present,which is also an important component of traffic informationization.In order to improve the application effect of license plate image recognition technology,this paper analyzes the classification and recognition technology of license plate image,improves the recognition efficiency on the basis of guaranteeing the recognition accuracy of license plate image.Firstly,this paper introduces the technology of license plate image classification and recognition,and understands the basic situation of this technology.Secondly,this paper introduces three common techniques of license plate image classification and recognition,which are license plate image collection technology,license plate image eigenvalue extraction and classifier,license plate image processing technology,to understand the application of different techniques in license plate image classification and recognition.Thirdly,three suggestions are put forward,which are to strengthen the collection of image content and quality control,to establish the database of license plate recognition samples,to define the license plate image recognition standard,and to clarify the development direction of license plate image classification and recognition technology in the future,in order to be able to future license plate image classification and recognition of the development of a solid foundation.

关 键 词：车牌图像 分类识别技术 卷积神经网络模型算法 sigmoid激活函数 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]