[1]邬忠萍,刘新厂,郝宗波.基于并行CNN和识别策略优化的车牌识别方法研究[J].南京师大学报(自然科学版),2023,46(03):98-104.[doi:10.3969/j.issn.1001-4616.2023.03.013]
　Wu Zhongping,Liu Xinchang,Hao Zongbo.Research of License Plate Recognition Method Based on Parallel CNN and Optimization Strategies[J].Journal of Nanjing Normal University(Natural Science Edition),2023,46(03):98-104.[doi:10.3969/j.issn.1001-4616.2023.03.013]
点击复制

基于并行CNN和识别策略优化的车牌识别方法研究()
分享到：

《南京师大学报（自然科学版）》[ISSN:1001-4616/CN:32-1239/N]

卷:
第46卷
期数:
2023年03期
页码:
98-104
栏目:
计算机科学与技术
出版日期:
2023-09-15

文章信息/Info

Title:
Research of License Plate Recognition Method Based on Parallel CNN and Optimization Strategies
文章编号:
1001-4616(2023)03-0098-07
作者:
邬忠萍1; 刘新厂1; 郝宗波2
(1.成都工业学院汽车与交通学院,四川 成都 611730)
(2.电子科技大学信息与软件工程学院,四川 成都 611731)
Author(s):
Wu Zhongping1; Liu Xinchang1; Hao Zongbo2
(1.School of Automobile and Transportation,Chengdu Institute of Technology,Chengdu 611730,China)
(2.School of Information and Software Engineering,University of Electronic Science and Technology of China,Chengdu 611731,China)
关键词:
车牌检测; 优化策略; 字符识别; 卷积神经网络; 智能交通系统
Keywords:
license plate detection;  optimization strategies;  character recognition;  convolutional neural network;  intelligent transportation system
分类号:
TP391
DOI:
10.3969/j.issn.1001-4616.2023.03.013
文献标志码:
A
摘要:
为改善车牌自动识别系统的通用性,在一般卷积神经网络(CNN)的基础上,提出一种具有两个浅层独立子网络的CNN,且具有并行卷积层计算的功能. 一个用于推理车牌的概率; 另一个利用线性激活对仿射参数进行回归. 支持对汽车(包括公交车和卡车等)、摩托车等不同类型的交通工具牌照的检测识别. 此外,使用基于YOLO v3的车牌字符识别模块,并施加了一系列的优化策略,实现对车牌中汉字的准确读取. 实验结果表明所提方法的识别精度优于一些同类优秀方法,在AOLP数据集上的车牌检测准确率达到98.9%,在CLPD数据集上的字符识别准确率达到96.2%. 所提方法有助于促进智能交通系统的进一步发展.
Abstract:
To improve the generalization of the automatic license plate recognition system,based on the general convolutional neural network(CNN),a CNN with two shallow independent sub-networks is proposed,which has the function of parallel convolutional calculation,one is used to infer the probability of license plate,and the other uses linear activation to regress affine parameters. The proposed method supports the detection and recognition of different types of vehicle license plates such as cars(including buses and trucks)and motorcycles. In addition,the license plate character recognition module based on YOLO v3 is used and a series of optimization strategies are applied to accurately read the Chinese characters in the license plate. The experimental results show that the recognition accuracy of the proposed method is better than other advanced methods,and the license plate detection on AOLP data set achieved a test accuracy of 98.9%,and the character recognition accuracy on CLPD data set is 96.2%. The proposed method is helpful to promote the further development of intelligent transportation system.

参考文献/References:

[1]ARAFAT M Y,KHAIRUDDIN A S M,KHAIRUDDIN U,et al. Systematic review on vehicular licence plate recognition framework in intelligent transport systems[J]. IET intelligent transport systems,2019,13(5):745-755.
[2]牛迪. 基于自注意力机制的多特征融合槽抽取模型[J]. 南京理工大学学报(自然科学版),2022,46(1):69-75.
[3]吴仁彪,冯晓赛,屈景怡,等. 雾霾环境下基于PLATE-YOLO的车牌检测方法[J]. 信号处理,2020,36(5):666-676.
[4]LI H,WANG P,SHEN C. Toward end-to-end car license plate detection and recognition with deep neural networks[J]. IEEE transactions on intelligent transportation systems,2018,20(3):1126-1136.
[5]XU Z,YANG W,MENG A,et al. Towards end-to-end license plate detection and recognition:A large dataset and baseline[C]//Proceedings of the European Conference on Computer Vision(ECCV). Munich,Germany,2018:255-271.
[6]KIM S G,JEON H G,KOO H I. Deep-learning-based license plate detection method using vehicle region extraction[J]. Electronics letters,2017,53(15):1034-1036.
[7]HSU G S,AMBIKAPATHI A M,CHUNG S L,et al. Robust license plate detection in the wild[C]//2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance(AVSS). Lecce,Italy:IEEE,2017:1-6.
[8]LAROCA R,ZANLORENSI L A,GONÇALVES G R,et al. An efficient and layout-independent automatic license plate recognition system based on the YOLO detector[J]. IET intelligent transport systems,2021,15(4):483-503.
[9]王昆,王晓峰,刘轩,郝潇.基于卷积神经网络的解扭曲车牌检测识别方法[J]. 计算机工程与设计,2021,42(11):3225-3231.
[10]李林,赵凯月,赵晓永,等. 基于卷积神经网络的污损遮挡号牌分类[J]. 计算机科学,2020,47(S1):213-219.
[11]SERGIO M S,CLAUDIO R J. License plate detection and recognition in unconstrained scenarios[C]//European Conference on Computer Vision. Munich,Germany:IEEE Press,2018:1-11.
[12]刘智. 基于改进自适应图像分割算法的车牌识别技术研究[J]. 西南师范大学学报(自然科学版),2017,42(5):28-33.
[13]XIANG H,ZHAO Y,YUAN Y,et al. Lightweight fully convolutional network for license plate detection[J]. Optik,2019,17(8):1185-1194.
[14]OMAR N,SENGUR A,AL-ALI S. Cascaded deep learning-based efficient approach for license plate detection and recognition[J]. Expert systems with applications,2020,149(12):113280-113291.
[15]段宾,符祥,江毅,等. 结合GAN的轻量级模糊车牌识别算法[J]. 中国图象图形学报,2020,25(9):1813-1824.
[16]XU H,GUO Z H,WANG D H,et al. 2D license plate recognition based on automatic perspective rectification[C]//2020 25th international conference on pattern recognition(ICPR). Milano,Italy:IEEE,2021:202-208.
[17]YAO Z J,YI W D. Bionic vision system and its application in license plate recognition[J]. Natural computing,2019,19(3):1-11.

备注/Memo

备注/Memo:
收稿日期:2023-02-15.
基金项目:国家自然科学基金资助项目(61003032)、轨道交通运维技术与装备四川省重点实验室开放课题基金项目(2020YW003)、成都工业学院引进人才科研启动项目(2021RC003)、四川省大学生创新创业项目(S202111116048).
通讯作者:邬忠萍,讲师,研究方向:人工智能应用,汽车智能化检测技术. E-mail:vitamin9991@163.com
更新日期/Last Update: 2023-09-15