Algorithm Research And Implementation Of License Plate Recognition System

License plate recognition system is the most critical research topic in the intelligent transportation system, it has a broad application prospects, such as road traffic monitoring, automated highway fees. As social and economic development, the number of motor vehicles is increasing,so the needs of vehicle safety management, guidance and control of traffic flow is more and more obvious, it is significance and useful to research and develop of more stable, fast and effective license plate recognition system.The key technology of license plate recognition is consisted of three parts:license plate location, character segmentation and character recognition. In this paper, apply image processing technology, pattern recognition, through deep studying and research of these three technologies, propose effective methods, and employ MATLAB platform, compile code to achieve license plate recognition system. Including the following:Propose a license plate location method based on line scanning and wavelet transform. Firstly, detect the edge of pretreatment image, and then locate the horizontal position of the license plate by line scanning method, left and right edges of the license plate can be figured out based on wavelet transform towards vertical projection of the horizontal area image as well, positioning a complete license plate.Propose a segmentation method based on license plate prior knowledge and vertical projection of character.The method combines inherent feature of the character such as the length, width of license plate and so on, roughly split the license plate by vertical projection method, and then make fine segmentation based on inherent feature of the license plate.After comprehensive analysis of the character recognition method based on template matching, propose a weighted template matching character recognition method by studying the lack of the traditional template matching, extract feature of pretreatment character, determine the result through calculating the distance between the character and the template, the characters that is difficultly distinguished is required of weighted matching.