Research Of Super-Resolution Restoration In Plate License Recognition

Plate license recognition is an important part of vehicles detection and traffic management with a variety of applications. The first step before recognition is to see Plate license image clearly and to enhance image quality, the recognition will have a very good effect. To see clearly license plate images, there is a method that is super-resolution restoration.For image degradation, this paper based on wavelet analysis and super-resolution and other image processing technology, doing a lot of researches in super-resolution restoration in license plate recognition from a low-resolution image, through analysis of existing theories and methods, has received the following new improved algorithm:(1) The general license plate image exists image degradation phenomenon, after pretreatment image could still be unable to identify or substantially reduce the recognition accuracy, so it is necessary to improve the image resolution in order to improve the image quality. Improvements in image spatial resolution depends primarily on super-resolution techniques, the paper on the various super-resolution methods were studied and compared, Presents a method based on wavelet domain of super-resolution improved algorithm for image restoration.(2) Super-resolution methods can get high-resolution images, through analysis of points motion estimation between the license plate image sequence, and the motion compensation we also can get super-resolution plate images. This paper presents super-resolution and motion estimation algorithm, the algorithm leads to a good result, through the low-resolution images motion estimation and the high-resolution images motion compensation.