Research On License Plate Recognition Technology On Intelligent Video Surveillance System

In early2003, China listed the first batch of pilot smart cities, which indicates that the development and planning of the city in China is transforming to a higher form of city from the original digital city to a higher form, smart city. Smart city makes the city more intelligent, high resource utilization, saving the cost of living, improving the quality of life and people-oriented continuous innovation by utilizing information and communication technologies. Smart city contains the six core systems which are connecting cooperating each other. Urban transport is one of the important parts of smart city. The research of China’s intelligent transportation system began in the late80s in last century. In recent years, remarkable results have been achieved, and then brought huge economic and social benefits to the community. Smart city has a great help to solve the city’s congestion, vehicle management, environmental pollution and conserve resources. However, the development of intelligent transportation systems in the various regions of China does not balance. Henan Province is in a slow development of it. Since there are six pilot of the smart cities in Henan Province, research and development of intelligent transportation systems related technology becomes even more urgent.In intelligent transportation systems, license plate recognition technology is an important part. This research subject comes from ShengLan Electronic Technology Co., Ltd., Zhengzhou City. The company is committed to independently research and traffic monitoring products. At present, the company has various products, such as adaptive control lights, electronic police, city traffic flow monitoring, regional control etc. Products. However, the company does not have a license plate recognition system.The main part of this paper is related to the algorithms for license plate recognition. After learning current license plate recognition technologies, the four major parts of the license plate recognition technology is implemented in our experiments. This paper also innovates the methods of recognition and location on license plate. (1) This paper summarizes the current pretreatment methods in license plate preprocessing, and uses two binaryzation methods according to the different characteristics of different binaryzation.(2) This paper uses the positioning method of rows and columns positioning on the different characteristics of the binary image. And it achieved good results.(3) This paper uses the method of combining vertical projection and template matching to segment characters. The method is simple and efficient with respect to the other methods.(4) For text and letters、numbers, this paper selects two different recognition methods to recognize characters. Outer contour characteristic algorithm, which has a higher recognition rate of the fuzzy characters, is used to identifying text; Template matching algorithm, which has a high recognition efficiency, is used to identifying letters and numbers.This paper finally describes the experimental platform, summaries the work, and presents our outlook on this subject.