License Plate Recognition Based On Neural Networks And Information Management Systems

As China's modernization process, road traffic has been high-speed development, traffic management automation and information technology has increasingly become a pressing problem, in this case, the automatic detection of vehicles as a source of information, obtaining more and more attention. On motor vehicle license and other related information automatically collected and processed for transport vehicles, park vehicles, parking management, traffic police inspector and so on great significance, as intelligent transportation is an important research topic. Intelligent Transportation has made in-depth study in research fields, and obtained a series of studies, most of intelligent traffic management point is the vehicle, the purpose is to make vehicles, scientific management, information technology and high efficiency.This article starts from license plate recognition, to do deep research of license plate recognition and license plate information management system based on the neural network, although existing previous research results, and have been used in practice, such as highway toll windows, parking, residential traffic management and so on, the license plate recognition rate did not reach 100%, the recognition rate of increase is still the focus of the study, in addition, the license plate management system used by local management, intranet data sharing, using the basic the client and server mode, but in a highly developed Internet browser today, browser and application server model is the mainstream.The research is based on residential vehicle management system, to realize the three parts that license plate recognition system unit, data storage unit and registration management system unit, using practical application to illustrate research results, the vehicle license's segmentation is realized by a new segmentation method improving the effect of the adhesive character, and providing a good identification environment for following characters. Character recognition use neural networks, and improve some parameters for a higher rate. Plate management system unit is using the most dominant browser and server mode, to achieve a cell management system used on a new architecture.In this issue, using VC++as the platform for license plate recognition algorithm, and using C++to realize precisely positioning plate, segmentation and neural network algorithm. Using SQL server2000 as the data-storing unit. Plate management system realized by JAVA implementation, operated or managed database through the B/S architecture network management system. The advantages of C language in image processing and that of JAVA development on the WEB were reflected. Making a good foundation for the latter function with expansion.