| With the development of economy and the improvement of standard of living,automobiles have stepped into household goods,becoming one of most important means of transportation. Although automobiles could bring to us a convenient and fast life,infrastructure construction can not keep pace with the increasing of vehicles, and a series of hidden danger has come out. The concept of intelligent transportation has been proposed.License plate recognition as important application of pattern recognition and computer vision,is a critical part of intelligent transportation system. It could be applied to transportation detection and management broadly to improve work efficiency and reduce administration cost.It contains two principle sections, license plate detection and plate character recognition.License plate detection, the key technology of license plate recognition, can not achieve the satisfactory level, so it is one of the destinations of this research area, with great research value and application prospects.Common image processing methods are introduced,including image smoothing, edge detection and binaryzation. A edge detection method is firstly proposed based on common detection methods and the character of Chinese license plate, which could highlight the characteristics of plate area, avoid other interference as much as possible at the same time.Then candidate regions are detected and other parts are removed rapidly based on statistical characteristic of edge which acquired by the proposed detection method. After that connected components of candidate regions are checked out by connected component analysis and selected by the cascade classifier to find out the character parts. The accurate locations of plates are determined based on relative positions of character connected component and the license plate template at last. Experimental result shows that the proposed method could achieve a good detection rate as well as detect multiple plates in a single image. |
PDF Full Text Request