An Ultra-Fast Automatic License Plate Recognition Approach for Unconstrained Scenarios

被引:8
|
作者
Ke, Xiao [1 ,2 ]
Zeng, Ganxiong [1 ,2 ]
Guo, Wenzhong [1 ,2 ]
机构
[1] Fuzhou Univ, Coll Comp & Data Sci, Fujian Prov Key Lab Networking Comp & Intelligent, Fuzhou 350116, Peoples R China
[2] Minist Educ, Key Lab Spatial Data Min & Informat Sharing, Fuzhou 350003, Peoples R China
基金
中国国家自然科学基金;
关键词
License plate recognition; Image recognition; Feature extraction; Object detection; Licenses; Character recognition; Task analysis; Automatic license plate recognition; multi-scale fusion; lightweight model; data augmentation; NEURAL-NETWORK;
D O I
10.1109/TITS.2023.3237581
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Recently, with the development of deep learning, Automatic License Plate Recognition (ALPR) has made great progress, However, there are still many challenges to accomplish license plate (LP) recognition under various traffic scenarios. One of them is the detection speed and recognition speed, and the other is the difficulty to recognize the low resolution and highly tilted LP images. In this paper, we present a two-stage ALPR framework to achieve efficient LP detection and recognition in unconstrained scenarios. Our LP detector is based on improved Yolov3-tiny, and we also propose a lightweight recognition network MRNet based on multi-scale features. In order to improve the inference speed, we abandoned the rectification of LP images, and RNNs that are difficult to compute in parallel. Additionally, we also propose a license plate data augmentation method, which achieves more effective augmentation and improves the generalization ability of the network through secondary random and hyperparametric search. As an additional contribution, we provide a challenging dataset collected from real-world driving recorders. The dataset is for multiple LPs in a single image, which makes up for the lack of public datasets in multiple LP scenarios. We evaluated the results on five datasets and showed that we achieved the best performance on almost all test sets, achieving 99.8% accuracy on CCPD with more than 180,000 license plate test sets. In terms of speed, the inference speed for detecting license plates reaches 751 FPS, and the fastest inference time for recognizing a single license plate takes only 2.9 ms.
引用
收藏
页码:5172 / 5185
页数:14
相关论文
共 50 条
  • [21] Pillar Number Plate Detection and Recognition in Unconstrained Scenarios
    Zheng, Shangdong
    Wu, Zebin
    Xu, Yang
    Wei, Zhihui
    Xu, Wei
    Liu, Jianxin
    Ding, Daohua
    Yang, Jiandong
    JOURNAL OF CIRCUITS SYSTEMS AND COMPUTERS, 2021, 30 (11)
  • [22] Automatic Recognition for Arbitrarily Tilted License Plate
    Lu, Nanxue
    Yang, Wei
    Meng, Ajin
    Xu, Zhenbo
    Huang, Huan
    Huang, Liusheng
    PROCEEDINGS OF 2018 THE 2ND INTERNATIONAL CONFERENCE ON VIDEO AND IMAGE PROCESSING (ICVIP 2018), 2018, : 23 - 28
  • [23] Automatic recognition of car license plate numbers
    Chen, MY
    Wong, EK
    PROCEEDINGS OF THE 8TH JOINT CONFERENCE ON INFORMATION SCIENCES, VOLS 1-3, 2005, : 745 - 748
  • [24] Research on Automatic License Plate Recognition Technology
    Guan Xuezhong
    Liu Shuang
    Li Zhi
    Li Yunjiang
    Song Taolue
    Liu Yang
    2013 THIRD INTERNATIONAL CONFERENCE ON INSTRUMENTATION & MEASUREMENT, COMPUTER, COMMUNICATION AND CONTROL (IMCCC), 2013, : 750 - 753
  • [25] Automatic License Plate Recognition Using Transformer
    Zhang, Teng
    Jia, Wei
    FOURTEENTH INTERNATIONAL CONFERENCE ON GRAPHICS AND IMAGE PROCESSING, ICGIP 2022, 2022, 12705
  • [26] Image simulation for automatic license plate recognition
    Bala, Raja
    Zhao, Yonghui
    Burry, Aaron
    Kozitsky, Vladimir
    Fillion, Claude
    Saunders, Craig
    Rodriguez-Serrano, Jose
    VISUAL INFORMATION PROCESSING AND COMMUNICATION III, 2012, 8305
  • [27] An effective automatic license plate recognition system
    Hu, PF
    Zhao, YN
    Zhu, JL
    Wang, JQ
    PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON IMAGING SCIENCE, SYSTEMS AND TECHNOLOGY, VOLS I AND II, 2001, : 101 - 106
  • [28] Skewness correction in automatic license plate recognition
    Huttunen, H
    IMAGE PROCESSING: ALGORITHMS AND SYSTEMS IV, 2005, 5672 : 130 - 138
  • [29] Automatic License Plate Recognition: A Comparative Study
    Ahmad, Imran Shafiq
    Boufama, Boubakeur
    Habashi, Pejman
    Anderson, William
    Elamsy, Tarik
    2015 IEEE INTERNATIONAL SYMPOSIUM ON SIGNAL PROCESSING AND INFORMATION TECHNOLOGY (ISSPIT), 2015, : 635 - 640
  • [30] A hybrid neuro-fuzzy approach for automatic vehicle license plate recognition
    Lee, HC
    Jong, CS
    APPLICATIONS AND SCIENCE OF COMPUTATIONAL INTELLIGENCE, 1998, 3390 : 159 - 168