| License plate recognition system is an important application and research in the field of intelligent transportation system using computer vision and pattern recognition. At present, some city in China develop their modern intelligent traffic control system, intelligent transport system can reducing traffic accidents,help the city set up a perfect traffic system. he license plate recognition system can monitor the condition about traffic automatically without any people. Some developed countries design the intelligent early,such as United States, Japan and European countries,they set up the system with the technology of sensors,camera,sonar and some other advanced technology.In intelligent image analysis,China is later than developed countries,so intelligent image analysis is very urgent in china now. therefore, this paper about the design about LPS is meaningful.With the help of Department of education of Anhui province natural science Fund,study the LPR system with landun company,and involved the R&D of LPR in bocoms company This paper introduced the license plate recognition system from four part,system architecture,hardware platform,algorithm of LPR and software design, in the hardware system which use the TMS320DM642 core introduced some main modules,including image acquisition module, memory expansion module and ethernet network interface, then introduced the algorithm of location,Segmentation of license plate,and character recognition;establish the algorithm with a large number of experiments,in the vehicle license plate location and angle adjust,use the global threshold algorithm and bersen algorithm which is local threshold binarizen the characters. the large number of experiments conducted to select the algorithm, at last introduced the program optimized and some attention problem in CCS.how to improve the efficiency of code is introduced in this part.software design section describes the software build environment CCS, DSP software development process, real-time DSP/BIOS, driving model, all the program is based on RF5 frame of reference.The system can capture and display image,the algorithm can realize the recognition of plate license, but it is still has some shortcomings.The system can not work in light or dark environments, With DM642 processor,it can develop the following function in this platform.the system has high practical. Supervised by LvHong... |
