Vehicle License Plate Recognition System, Key Technology Research

Being a special computer vision system, the Vehicle License Plate Recognition (VLPR) System can capture a vehicle image automatically and identify the vehicle license plate (VLP) in the image. The VLPR system involves the techniques of digital image processing, computer vision, pattern recognition and artificial intelligence and so on. Its key techniques include the VLP location, character segmentation and character recognition and so on. We studied on the three key techniques of the VLPR system. The VLP location, which is to locate the VLP from a vehicle image, is the base of the VLPR system. This paper proposed an algorithm of VLP location aimed to a complex background. This algorithm adopted manifold methods, such as edge detection, component connectivity analysis and slant correction, to resolve the problem of locating license plate in a complex background. This method has the capability of accurately locating license plate in a complex background, and is insensitive to weather, illumination variation, motion and the slant of the license plate, It also provides the slant angle, character orientation etc information for the succeeding work. The VLP character segmentation means to segment the license image into single character for single character recognition. We adopt connectivity-based character segmentation algorithm, rather than the traditional projection based one, and some elementary results are given. Chinese characters, letters and numbers in the license plate are...