The Design And Implementation Of Car License Plate Automatic Recognition System

This paper mainly studies the vehicle license plate automatic recognition system, as an information, intelligence, digital integrated traffic management system is very important in today’s automobile universal situations. Due to the diversification of the license plate in our country, different style, color, specifications, this paper mainly aimed at the plate size is 440 mm * 140 mm specifications, the size of this licence can account for over 90% of the license plate in daily life. Based on digital image processing, pattern recognition and computer vision technology, with their knowledge of license plate recognition and research, in the process of system implementation, will determine the framework for image preprocessing, license plate locating, character segmentation and character recognition phase, the implementation of every stage, from the perspective of the traditional theory, combined with experiments and experience to add a new algorithm and the train of thought, to improve on efficiency. In the image preprocessing stage, using the image gray level to reduce the redundant information of source image, make take up very little space, the use of image enhancement increases the thread of the original image is not very clear, clear vehicle information is presented. In license plate locating phase, first analyzed the characteristics of the license plate information, under the feature information and prior knowledge find the connected area of the license plate, then on the basis of the analysis of the texture information, such as remove false license plate area, finally find the license plate area. In character segmentation stage, the first level of license plate tilt correction, vertical and horizontal projection of license plate frame and rivets, etc., and on the vertical projection, appear regularly in the area of the projection interval, judgment and analysis these intervals, finally to determine the left and right boundaries of each character, character segmentation. Finally, character recognition stage, at this stage with the improved template matching algorithm to partition the individual characters of license plate recognition. The entire process of system implementation, in vs2008 development environment, completed the development of the whole system in c + + language. Closely linked to each stage, on the basis of this the car license plate automatic recognition system modules design. Finally, to complete the system integration and operation, and through the test, the system recognition rate has greatly improved.