| Due to the rapid increase in the number of motor vehicles,the general traffic management mode cannot satisfy the demand for the management and flow regulation of existing vehicles.The concept of intelligent transportation system is proposed to establish a more intelligent,fast and automated transportation system to alleviate traffic congestion.One of the core technologies of intelligent transportation is the license plate recognition technology,the research of license plate recognition technology provides a firm technical foundation for the establishment of intelligent transportation system.The positioning of the license plate region and the recognition of license plate characters are the two core parts of the license plate recognition system,which determines the performance of the whole system.This paper designs and implements the vehicle license plate recognition system based on image energy and neural network,and completes the design of the simple interface of the system through the comprehensive study and comparison,after extensive reading and research of the document of license plate recognition system.Through the research and summarization of the common license plate region localization algorithm,the algorithm of vehicle license plate locating based on image energy is realized and verified in the experiment.The algorithm firstly constructs the energy image of the preprocessing gray plate image,then locates the three candidate regions of the energy image in the first place,finally screening and verifying the candidate region of the initial positioning,completing the precise positioning of the plate area and obtaining the license plate area.After the in-depth study and comparison of the commonly used license plate character recognition algorithms,the method of two-level BP neural network is used as the character recognition method of the license plate recognition system.The algorithm improves the character feature extraction method and adopts the two level BP neural network:The first level includes the character network and the digital alphabet network,the second level is similar character network.Many experiments show that the character recognition rate can be improved greatly by constructing similar character network.This system is based on the QT development platform to design the license plate recognition system GUI interface,which has certain operability and the cross-platform characteristics of different hardware platforms.The experimental results show that the vehicle license plate recognition software system can identify the image of the license plate in different angles,different distances and different illumination environment.Due to the introduction of QT,the software is convenient to operate on Win+PC machine or Linux+ARM system,with strong application foreground. |
