Malaysian vehicle plate number identification using deep learning.

Automatic License Plate Recognition (ALPR) is a useful tool for preventing unlicensed, forbidden, or uninsured drivers, as well as auto theft. Many practical applications, such as traffic law enforcement, road traffic monitoring, automatic toll collection, and private area access management, have made it a popular research topic. Due to various intrinsic elements such as variability in character pattern, font, shape, size, colour, and plate dimensions, obtaining and identifying the license plate region in the entire traffic image can be extremely difficult. On another note, Deep Learning has a significant history of success, outperforming humans in image recognition, speech recognition, translation, and a variety of other tasks. Thus, in this study, Deep Learning was employed to provide an automatic Malaysian vehicle plate number identification. The YOLOv3 was used to train the data at 5000 epochs. On the other hand, the TensorFlow was employed to detect the plate number region, while the Tesseract OCT was utilized to recognize the plate number characters. The application of Deep Learning to 200 testing images returned 100% accuracy of plate number region detection, and 84.17% of overall mean percentage of character recognition (CR) accuracy which indicating good overall detection accuracy. It can be inferred that the proposed implementation of Deep Learning for Malaysian vehicle plate number identification is found to be successful. [ABSTRACT FROM AUTHOR]

Copyright of AIP Conference Proceedings is the property of American Institute of Physics and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)