License Plate Detection and Character Recognition Using Deep Learning and Font Evaluation
Zahra Ebrahimi Vargoorani, Ching Yee Suen
TL;DR
This work tackles the sensitivity of license plate recognition to lighting and font diversity by proposing a dual-stage framework that couples Faster R-CNN-based license plate detection with a CNN–RNN OCR trained with $CTC$ loss, using a lightweight $MobileNetV3$-style backbone for edge efficiency. It leverages the UFPR-ALPR and CENPARMI datasets, augmented with synthetic data, to evaluate both detection and recognition across fonts such as Driver Gothic, Dreadnought, California Clarendon, and Zurich Extra Condensed. The results show recall of approximately $92\%$ on CENPARMI and $90\%$ on UFPR-ALPR, with significant improvements over a OpenALPR baseline and clear font-dependent performance variations demonstrated via confusion matrices. The study highlights the practical impact of font design on ALPR and provides guidance for font-aware system improvements and future dataset/architecture enhancements for more robust, real-time deployment $on$ edge devices.
Abstract
License plate detection (LPD) is essential for traffic management, vehicle tracking, and law enforcement but faces challenges like variable lighting and diverse font types, impacting accuracy. Traditionally reliant on image processing and machine learning, the field is now shifting towards deep learning for its robust performance in various conditions. Current methods, however, often require tailoring to specific regional datasets. This paper proposes a dual deep learning strategy using a Faster R-CNN for detection and a CNN-RNN model with Connectionist Temporal Classification (CTC) loss and a MobileNet V3 backbone for recognition. This approach aims to improve model performance using datasets from Ontario, Quebec, California, and New York State, achieving a recall rate of 92% on the Centre for Pattern Recognition and Machine Intelligence (CENPARMI) dataset and 90% on the UFPR-ALPR dataset. It includes a detailed error analysis to identify the causes of false positives. Additionally, the research examines the role of font features in license plate (LP) recognition, analyzing fonts like Driver Gothic, Dreadnought, California Clarendon, and Zurich Extra Condensed with the OpenALPR system. It discovers significant performance discrepancies influenced by font characteristics, offering insights for future LPD system enhancements. Keywords: Deep Learning, License Plate, Font Evaluation