Research Into Vehicle License Plate Recognition System

Being applied in areas like toll payment and traffic management, the license plate recognition technology has the functions of reducing manpower, increasing efficiency and improving management system, etc. As Intelligent Transport Systems are developing rapidly in China, the license plate recognition technologies show great importance both economically and realistically.Some license plate recognition products have been put into market in China. However they can only achieve a high recognition rate under an environment of enough light, and the license plate also must be clean enough. Therefore, the flexibility is not good enough according to the practical work circumstance. This thesis is based on insight research on the license plate recognition technology, and deals mainly with the illegibility of license plates that arises from bad weather, insufficient light and awful stain condition. We also try to make the algorithm more adaptable to the environment.The thesis makes systemic research on three steps included in the algorithm. They are license plate location, character segmentation and character recognition. For the plate location problem,we use a detection method based on one derivative arithmetic operator, which can focus on the texture in license plate area. During the location course we propose the new double-smoothing method, which makes the projection image smooth and consecutive, as well as keeps the original character furthest.For the character segmentation, depending on the traditional vertical projection segmentation and the transcendental width and height knowledge of every character, we propose a projective characteristic value and a character segmentation method based on this projective characteristic value. This method achieves a good performance in correcting errors and eliminating interference.