PatrolVision: Automated License Plate Recognition in the wild
Anmol Singhal Navya Singhal
TL;DR
The paper tackles real-time automated license plate recognition (ALPR) in unconstrained urban patrolling by introducing a Singapore-focused dataset and an end-to-end, edge-deployed ALPR prototype. It combines a lightweight LPD stage using RFBNet with a YOLO-based LPR stage to detect and recognize characters, followed by post-processing heuristics and a robust character arrangement module to handle both single-line and double-line plates. On a dataset of 16,910 frames, the system achieves 86.1% plate detection precision and 67% full-character recognition accuracy, with higher partial-match scores (89%/95%), and demonstrates real-time performance up to around 64–68 FPS on modern GPUs across platforms. The work highlights practical viability for patrol vehicles using dashcams and edge GPUs, while suggesting future improvements via alternative recognizers and domain-specific post-processing to further boost accuracy.
Abstract
Adoption of AI driven techniques in public services remains low due to challenges related to accuracy and speed of information at population scale. Computer vision techniques for traffic monitoring have not gained much popularity despite their relative strength in areas such as autonomous driving. Despite large number of academic methods for Automatic License Plate Recognition (ALPR) systems, very few provide an end to end solution for patrolling in the city. This paper presents a novel prototype for a low power GPU based patrolling system to be deployed in an urban environment on surveillance vehicles for automated vehicle detection, recognition and tracking. In this work, we propose a complete ALPR system for Singapore license plates having both single and double line creating our own YOLO based network. We focus on unconstrained capture scenarios as would be the case in real world application, where the license plate (LP) might be considerably distorted due to oblique views. In this work, we first detect the license plate from the full image using RFB-Net and rectify multiple distorted license plates in a single image. After that, the detected license plate image is fed to our network for character recognition. We evaluate the performance of our proposed system on a newly built dataset covering more than 16,000 images. The system was able to correctly detect license plates with 86\% precision and recognize characters of a license plate in 67\% of the test set, and 89\% accuracy with one incorrect character (partial match). We also test latency of our system and achieve 64FPS on Tesla P4 GPU