Research On License Plate Location Technology Based On Edge Point Detection

Due to the continuous development of economy and society in recent years, the number of motor vehicles has increased. This has caused great pressure on the traffic management. In order to solve this problem, many scholars have studied different intelligent transportation systems which can be used to manage the traffic conditions.The License Plate Recognition system is widely used in this field. This paper mainly studies a key step in the LPR system: License Plate Location.In this paper, firstly, some current technologies of License Plate Location have been studied. Then, I analyze and discuss some new methods of License Plate Location in recent years. Finally, the License Plate Location method in this paper is presented. In the pre-treatment stage, firstly, the color image is processed by the weighted average method. Then I use the median filtering to smooth the gray image,and remove some noise in the image. In the stage of initial location, the Sobel Edge Detection is used to process the pre-processed image. The Binary Image is obtained by using the algorithm of Otsu. A rectangular sliding window is constructed to generate an edge density matrix. Then, the expansion program of CEP is used to remove a lot of non-plate area. So that the data of image is greatly reduced. In the stage of precise location, based on the technique of left-up edge-point detection, an evaluation function is constructed by using the edge features of the license plate area.Then each detected edge points is evaluated. And the point of satisfying the condition is selected. Finally the precise location of the license plate is completedIn this paper, one of the existing methods is improved. And the time and accuracy of the experiment have been improved. In this paper I simulate 283 images by using the M language program of MATLAB. The results of the experiments show that the algorithm is effective.