车牌识别系统算法的研究电子科学与技术电路与系统专业论文.docxVIP

兰州大学硕士学位论文Abstract 兰州大学硕士学位论文 Abstract Intelligent transport system has become a focus of research attention．License Plate Recognition System is the main field of intelligent transport system，which plays an important role in intelligent traffic control system，parking lot monitor system and automatic charging system． License plate recognition system consists of image pre—processing，vehicle license plate location，character segmentation and character recognition and other major modules．Based on the research,analysis of existing algorithms based on the theory of image preprocessing and character recognition algorithm has been improved and perfected．The most distinctive parts would be described in following aspects： 1．In the image pre—processing module of this paper,by analyzing the color,size and gray level distribution of geometric features of license plate area，an integrated method based on the using of the edge of the color pair，morphological integration—corrosion and PCNN binarization is proposed．，nle method Can get rid of noise edge and achieve the purpose of outstanding license plate information． 2．License plate location can be divided into coarse positioning and fme positioning， rough location of license plate area is determined by statistics of edge characteristics． Fine border of license plate is gotten in the use of color features and horizontal projection of binary image． 3．In the character segmentation module，after vertical tilt adjustment，use the vertical projection to split characters，and at the same time，we use the middle line to segment adhesion characters． 4．In accordance谢m features of characters of license plate，matching degree is calculated by improved Hausdorff distance and similar characters are processed in secondary breakdown to improve the character recognition rate． This algorithm can cater to license tilt，distortion，character blurred，light effect．After experimental verification，based on this algorithm license plate recognition system Can achiev