Abstract
With the rapid rise of private vehicles around the world, License Plate Recognition (LPR) plays a vital role in supporting the government to manage vehicles effectively. However, an introduction of new types of license plate (LP) or slight changes in the LP format can break previous LPR systems, as they fail to recognize the LP. Moreover, the LPR system is extremely sensitive to the conditions of the surrounding environment. Thus, this paper introduces a novel deep learning-based Korean LPR system that can effectively deal with existing challenges. The main contributions of this study include (1) a robust LPR system with the integration of three pre-processing techniques (defogging, low-light enhancement, and super-resolution) that can effectively recognize the LP under various conditions, (2) the establishment of two original Korean LPR approaches for different scenarios, including whole license plate recognition (W-LPR) and single-character license plate recognition (SC-LPR), and (3) the introduction of two Korean LPR datasets (synthetic data and real data) involving a new type of LP introduced by the Korean government. Through several experiments, the proposed LPR framework achieved the highest recognition accuracy of 98.94%.
Highlights
As one of the convenient and comfortable means of transportation, the number of motor vehicles has rocketed in recent years
A total of 510,500 license plate (LP) images were obtained for the W-license plate recognition (LPR) method, of which 500,000 synthetic images were generated by a python script, and 10,500 real images were taken in different scenes. 216,000 images consisting of various characters were generated to feed the model of the single-character license plate recognition (SC-LPR) method by using the proposed character segmentation method
A model combining CNN and RNN was used to train the images of whole LPs in the whole license plate recognition (W-LPR) method
Summary
As one of the convenient and comfortable means of transportation, the number of motor vehicles has rocketed in recent years. As the license plate (LP) is the only way to identify a vehicle, automatic license plate recognition (LPR) systems have brought significant changes to traditional vehicle management. The common LP format of private vehicles is made up of two initial digits, one Korean character, followed by four digits. There are three common processes for the LPR, including license plate detection (LPD), character segmentation, and recognition. A novel LPR system with a robust performance was proposed in this study, which integrates computer vision (CV), image processing, and deep neural network technology. A large dataset is required to achieve good recognition accuracy by using deep learning methods, but the data of various Korean characters are not easy to be obtained. 4. The collection of a large Korean license plate dataset that contains both real LP images and synthetic LP images.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
