LPTR-AFLNet: Lightweight Integrated Chinese License Plate Rectification and Recognition Network

TL;DR

LPTR-AFLNet is a lightweight, end-to-end network that effectively rectifies perspective distortions and recognizes Chinese license plates in real-time, suitable for deployment on resource-constrained edge devices.

Contribution

The paper introduces a novel unified network combining correction and recognition modules, with innovative guidance and improved recognition components for Chinese license plates.

Findings

High accuracy in perspective distortion correction and recognition.

Real-time processing speed under 10 milliseconds on mid-range GPUs.

Effective handling of complex and challenging license plate scenarios.

Abstract

Chinese License Plate Recognition (CLPR) faces numerous challenges in unconstrained and complex environments, particularly due to perspective distortions caused by various shooting angles and the correction of single-line and double-line license plates. Considering the limited computational resources of edge devices, developing a low-complexity, end-to-end integrated network for both correction and recognition is essential for achieving real-time and efficient deployment. In this work, we propose a lightweight, unified network named LPTR-AFLNet for correcting and recognizing Chinese license plates, which combines a perspective transformation correction module (PTR) with an optimized license plate recognition network, AFLNet. The network leverages the recognition output as a weak supervisory signal to effectively guide the correction process, ensuring accurate perspective distortion correction. To enhance recognition accuracy, we introduce several improvements to LPRNet, including an improved attention module to reduce confusion among similar characters and the use of Focal Loss to address class imbalance during training. Experimental results demonstrate the exceptional performance of LPTR-AFLNet in rectifying perspective distortion and recognizing double-line license plate images, maintaining high recognition accuracy across various challenging scenarios. Moreover, on lower-mid-range GPUs platform, the method runs in less than 10 milliseconds, indicating its practical efficiency and broad applicability.