Deep-Learning-Based Approach for Iraqi and Malaysian Vehicle License Plate Recognition.

Recognizing vehicle plate numbers is a key step towards implementing the legislation on traffic and reducing the number of daily traffic accidents. Although machine learning has advanced considerably, the recognition of license plates remains an obstacle, particularly in countries whose plate numbers are written in different languages or blended with Latin alphabets. This paper introduces a recognition system for Arabic and Latin alphabet license plates using a deep-learning-based approach in conjugation with data collected from two specific countries: Iraq and Malaysia.