Research on Deep Learning-Based Vehicle Number Plate Recognition Systems

View at publisher Summarise with VIRAT
Authors: Rumeng Jia
Cited by: 0
References: 8

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This paper addresses the challenges in vehicle licence plate recognition tasks within complex environments by proposing a deep learning-based approach incorporating an enhanced YOLOv8n model and a Convolutional Recurrent Neural Network (CRNN). Firstly,the CBAM attention mechanism is integrated into the YOLOv8n object detection model to enhance feature extraction capabilities within licence plate regions, thereby improving detection accuracy and robustness. Secondly, a bidirectional LSTM (BiLSTM) is incorporated into the CRNN architecture to better capture contextual dependencies between characters within the licence plate, thereby improving character recognition accuracy. Experiments conducted on the CCPD dataset demonstrate that the proposed algorithm outperforms mainstream detection models, exhibiting excellent real-time performance and environmental adaptability. It effectively handles complex scenarios such as varying illumination, weather interference, and angular inclination, providing a reliable solution for licence plate recognition tasks within intelligent transport systems.

Fields of Study & Subjects

Vehicle License Plate Recognition Advanced Neural Network Applications Handwritten Text Recognition Techniques

Publication details

Journal / source
IEEE International Conference on Power, Electronics and Green Energy (ICPEGE)
Pages
1051–1055
Publication date
28 January 2026
Language
EN