A Study Of Automatic Orientation Of Car License Plate Targets

With the sharp increase of the motor vehicle amounts, the intelligent traffic management is highly demanded more and more. However, the critical technology that can implement the intelligent traffic control, the technology of car license plate targets orientation, is at elementary phase and needs further research.First of all, the technical route applied in the license plate orientation system, and the methods to solve the key problem are discussed in this dissertation, which involves technology about image processing and identification. In the second place, the whole structure of license plate orientation is studied in this dissertation: at first, the car photograph should be processed by smooth filtering, gray stretching, binarization, then the location of the car target can be segmentated from the complex background based on the changes characteristic of the word. And some new algorithms and methods are presented in the dissertation such as: pre-processing of image, edge detection, license plate orientation and segmentation.In order to improve the efficient of the license plate location system, the mothod named "Adaptive threshold and dynamic analysis with a variable step" based on the characteristic of the gray and the texture of the car target is presented in the dissertation which can improve the orientation speed; The method named "Automatic orientation of car license plate targets based on the unvariable characteristic of the projection" is proposed in the dissertation in the terms of the imaging characteristics of the car license plate target. It can succeed in locating the car target from the complex background when the angle of inclination is relatively small. " A new way based on Hough transform to locate the edge of the car target" was proposed and was also applied which can get good location result and can meet the aim of the study of the project. Finally, the problems and the future directions of the car target orientation are discussed.