License Plate Recognition System And The Research Of Its SOPC Technology Implementation

With the rapid development of information, the recognition technology of vehiclelicense plate has become an important component of China’s intelligent transportation and hasbeen widely used for car park management, highway electronic toll system ETC (ElectronicToll Collection) and illegal traffic management system. It has great significance for buildingintelligent transportation systems (ITS). Through the study of the license plate recognitionalgorithm, this paper presents a new algorithm. And did some research of the algorithmimplementation by SOPC technology. Finally the article designed a license platerecognition system with the performance of high-speed, accurate, low power.In the side of software design of system, the paper has made some improvements for theshortcomings on the vehicle license plate location, character segmentation and characterrecognition. The positioning of the license plate license plate recognition based on the abilityto accurately locate the license plate relationship to the success or failure of the entire system,the license plate location, the license plate location method based on edge detection, edgedetection and morphological operations to achieve the positioning of the license plate;character segmentation is the bridge of the license plate location and character recognition, Inthis paper, the projection algorithm to achieve a good segmentation of the characters;algorithm of license plate character directly affect the recognition speed of the system and thefinal results, this paper in the neural network character recognition technology has beenimproved and a large number of experiments, the full application of SOPC technologymodules run in parallel features, characters and numbers, letters, while using differentnetworks in parallel to identify and propose a sub-region weights of similar character for thelicense plate in the form proofreading, and recognition speed and accuracy has improved tosome extent.In the side of system hardware, the thesis combines the vehicle license plate recognitiontechnology with SOPC (System on the Programmable Chip System on Programmable Chip)technology. License plate recognition will be counted in the Nios II processor is built fortransplant, and part of the algorithm is simple and suitable for parallel processing operation described by the Verilog HDL language, application hardware acceleration system to increasethe processing speed. Try using software, hardware a combination of methods to achieve anew way of vehicle license plate recognition algorithm, the vehicle license plate recognitiontechnology and SOPC technology, refinement and commercialization played a certain role.