The Research On The Certain New Key Algorithms Of License Plate Recognition System

With the development of information technology and intelligence technology, the informalization and intelligentizing of traffic management has already been the trend. License Plate Recognition system (LPR) is the core of the Intelligent Traffic System (ITS). And it plays an important role in modern traffic management systems. LPR is the technique to get the character information from vehicle image and confirm the identity of the vehicle through computer vision, image manipulation, and pattern recognition and so on. LPR system has great value of application in the common security, traffic management, flow observation, electronic charge system and related military department.In this paper, three key technologies including license plate location, character isolation and character recognition are discussed. License plate location is basis and prerequisite for the entire LPR system and directly relate to smoothly proceeding of the next two processes (Character isolation and Character recognition). However, due to license plate area in the whole digital image is not fixed in position and the color and size of license plates are not sure and the deviation brought about by light and complex environment ,so fast and accurate positioning of license plate region have certain degree of difficulty. A Method for the Multi-threshold segmentation and Location of License plate Based on HSV Color system is proposed in this paper. Since there are a lot of skew plates in the images and in order to get accurate plate region, so after the plate location the skew angel need to be calculated and adjusted.Character isolation is a process that gets a series of individual characters from the plate region. By using projection method to find border of the characters, and in accordance with the known characters of license plates with geometric knowledge and rules to solve these problems, so precise boundaries of the characters image can be achieved. Character is the core of LRP system. By using the structure of the character recognition method, extracting the strokes from the normalization character images and then designing a multi-level tree identification, characters can be correctly identified.The result of experiments shows that the algorithms used in this paper have high rate of plate recognition. And the executing speed is fast. The algorithms can be applied to actual projects. The system we implement in this paper is a valid LPR system.