Lightweight and accurate license plate reader designed for police officer's cell phone

Nowadays, mobile phones have provided the opportunity to collect information and process it, in this regard, the use of mobile phones for police officers can also be very effective. In this article, it is suggested that police officers have a mobile phone in their pocket equipped with a license plate reader program to record license plate data even in normal patrols. This article provides a lightweight and accurate license plate reading method suitable for mobile devices that can be implemented in real time. In the proposed method, morphological and binary methods are used to locate the license plate, then the character segmentation is performed using the real-time thresholding method. Finally, by selecting a quick classifier and training it based on a large number of Iranian license plate character patterns, the characters are classified. The main purpose of this article has been to expose the speed of execution and accuracy with the aim of providing the possibility of real-time implementation on mobile. Due to the fact that the existing articles do not provide resources commensurate with the implementation conditions on mobile, the results of the proposed method are compared with existing real-time methods that are implemented without the need for a GPU. The simulation results show that the proposed method has the ability to read the license plate accurately along with the ability to implement it on a mobile phone almost in real time.