Research And Application Of The Vehicle License Plate Recognition(VLPR) System

Being a specia1 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 artificia1 intel1igence and soon on. Its key techniques include the VLP location,character segmentation and character recognition.We studied on the four key techniques of the VLPR system. The thesis firstly translate multicolor image into gray image to faster the speed of image processing, then eliminate the noises. After that,to image enhancement through counterchange gray or filtering. The VLP location,which is to 1ocate the VLP from a vehicle image,is based on geometry feature location. Firstly, we do fringe detection with vehicle image. Secondly, conceivable area of VLP will be chose. At the last, we ensure which is the real VLP area. The VLP character segmentation means to segment the 1icense image into sing1e character for sing1e character recognition .We adopt a technique of character segmentation integrated by connectivity-based character segmentation algorithm, projection and clearance character segmentation algorithm, rather than the traditional projection based one,and some elementary results are given. Chinese characters, 1etters and numbers in the 1icense plate are perceived in the character recognition process that is the kernel part of LPR system. The thesis introduce Integration Invariant Moments (IIM) on which we do some research and validation to extract features. IIM this thesis mentioned has stronger capabilities of rejecting noises and interferences. It can extract invariants of blurred target in complex background effectively, enhancing the reliabilities of the movement blurred target recognition enormously.The thesis, which is integrated by the methods mentioned all above, is adopted modularization design to VLPR System and implement a system which is provided with integrated function of recognition.