🤖 AI Summary
This study addresses the challenges of license plate detection and recognition (LPDR) in data-scarce regions with visually distinctive plates, such as Bolivia, where performance is often degraded by perspective distortion and illumination variations. The authors propose a two-stage robust recognition framework: first, a YOLO-based detector is pre-trained on synthetic data generated via Blender to simulate extreme imaging conditions, followed by domain-adaptive fine-tuning using street-view imagery from La Paz. Detected plates undergo geometric rectification before being fed into a character recognition model, with a lightweight vision-language model (Gemma3-4B) serving as a fallback mechanism triggered by low prediction confidence. This work introduces the first public LPDR dataset for Bolivian license plates and achieves a character-level accuracy of 89.6% on real-world street scenes, significantly enhancing recognition robustness in complex urban environments.
📝 Abstract
Robust license plate recognition in unconstrained environments remains a significant challenge, particularly in underrepresented regions with limited data availability and unique visual characteristics, such as Bolivia. Recognition accuracy in real-world conditions is often degraded by factors such as illumination changes and viewpoint distortion. To address these challenges, we introduce BLPR, a novel deep learning-based License Plate Detection and Recognition (LPDR) framework specifically designed for Bolivian license plates. The proposed system follows a two-stage pipeline where a YOLO-based detector is pretrained on synthetic data generated in Blender to simulate extreme perspectives and lighting conditions, and subsequently fine-tuned on street-level data collected in La Paz, Bolivia. Detected plates are geometrically rectified and passed to a character recognition model. To improve robustness under ambiguous scenarios, a lightweight vision-language model (Gemma3 4B) is selectively triggered as a confidence-based fallback mechanism. The proposed framework further leverages synthetic-to-real domain adaptation to improve robustness under diverse real-world conditions. We also introduce the first publicly available Bolivian LPDR dataset, enabling evaluation under diverse viewpoint and illumination conditions. The system achieves a character-level recognition accuracy of 89.6% on real-world data, demonstrating its effectiveness for deployment in challenging urban environments. Our project is publicly available at https://github.com/EdwinTSalcedo/BLPR.