License Plate Location And Identification Of Key Algorithm

With increasing motor vehicles, how to achieve automation and intelligent road traffic management, and enhance road safety, road congestion inhibition to ensure the transport efficiency, so it has a convenient, comfortable and safe traffic environment has become a serious problem in society today. License Plate Recognition is one of the fundamental technologies in Intelligent Transportation System, mainly through the license plate character recognition means to identify vehicle identification, license plate location is the key technical in license plate recognition system, if not exactly looking for images of vehicles to license plate location, follow-up of license plate character segmentation and recognition of the work can not proceed.This paper presents the popular domestic and foreign vehicle license location identification technology, mainly on the key algorithm of the license plate location and recognition is studied and discussed.First, concerning the license plate location problem computationally intensive and positioning a long time, a fast and practical positioning method is advanced. The main idea is to first on the license plate image histogram equalization pre-processing, and candidate district is determined with improved Prewitt edge detection operator and Projection and Mathematical Morphology. Then, the interfered district is excluded by the feature of geometric and color. The final positioning of the various environments the results are compared, and gives the corresponding experimental results.The second part of the license plate location is mainly explained later, the follow-up in the LPR system, two areas of work, this paper analyses thoroughly the license plate character segmentation, identifying links the characteristics of each of the components, as well as achieving the link commonly used method, Hough transform main advantage of the change on the location of the plate to tilt after the correction, projection of character segmentation, template matching pairs of license plate character recognition, and through experiments realized Segmentation and Recognition of license plate character.That the proposed fast location method for the multi-plate, the application in real life found that the vehicle is suitable for residential applications such as vehicle management, and underground parking areas, which have some of the value and social and economic benefits.