The Study Of Key Technology For License Plate Recognition

With the increase of motor vehicles,how to achieve automation and intelligent road traffic management has become a serious problem in society today. Vehicle license plate recognition technology is an important research topic in intelligent traffic system.It has a wide range application in urban traffic control system,traffic flow station,automatically record traffic violations,highway bayonet charges,intelligent park management and other traffic automation system.Therefore,the research and development of the technology is of great practical value and real significance.This paper is mainly about the key algorithm of the license plate recognition. It mainly includes four parts of image preprocessing, license plate location, character segmentation and character recognition. It improves the edge detection algorithms and optimizes the existing algorithm. The text introduces an improved license plate location algorithm based on texture characteristics , it improve the speed of location of the license plate. It is proved that the method can locate the home license plate clearly and accurately, but also can accurately determine the plate position for foreign vehicles in the matlab environment simulation. In the character segmentation, it achieves the character segmentation accurately using the projection threshold which is able to change according to different images. Experimental results prove that this method can overcome the phenomenon of over segmentation and character adhesion. Finally the improved BP neural network is used to identify each character.Experimental results show that the method can achieve accurate license plate recognition accurately and it has the characteristics of accurate positioning, robustness and so on.