Research And Application Of License Plate Recognition And Traffic Detection Algorithm In Intelligent Transportation System

With the improvement of living standards, private car ownership increases substantially. The automobile popularization, on the one hand, enriches the people’s daily travel; on the other hand, gives the load capacity of the road traffic system a great challenge. Intelligent transportation system (ITS) points out the future development direction of the traditional traffic system. The license plate recognition technology and violation detection are important parts of intelligent traffic management system. This paper mainly studies the license plate recognition and violation detection technology, the specific contents are as follows:(1) License plate location. The paper studies the license plate location algorithm based on edge detection and jump point. With the license plate location, this paper presents an algorithm to look for car body area, and then makes the common body color identification.(2) Character segmentation and recognition. Based on the survey data, the plate is segmented with projection and prior knowledge of license plate, the algorithm is robust to character conglutination and fracture. In character recognition part, the character feature is extracted with existing grid method, and then support vector machine is used to train and distinguish character samples.(3) Violation detection. This paper mainly studies the over-speed detection. Using mixed gauss background modeling, it realizes the car speeding detection algorithm based on video. The field test shows that this method is simple and fast.Finally, the license plate recognition and illegal detection algorithm are simulated.