Research For License Plate Recognition System Based On Compressed Sensing

With the development of the modern electronics, computer, and automatic control, intelligent transportation system get the rapid development platform, absolutely including the importance of license plate recognition system in modern traffic. For example, monitoring some speeding or parking vehicles. License plate recognition is a key for it. Therefore, it is a necessary and very important step to pay attention to license plate recognition in intelligent transportation system, many methods have been applied to build license plate recognition system. At the same time, the license plate recognition technology developed more and more quickly. License plate recognition system is also becoming an important research emphasis in the academic community.Firstly, the article make a brief introduction about the process of license plate recognition system, describing the theoretical framework of Compressed Sensing as well as its application areas. The main work focus on researching the digital image processing system in the process from image acquisition device to main system, Compressed Sensing theory would be applied to the process of the license plate digital image processing in system.Then, compared with the currently algorithms, their time complexity or reconstruction effect is ineffective. Above algorithm have some problems, so we proposed a new type of license plate digital image processing method based on Compressed Sensing. Then, do some entire process for license plate digital image after improve wavelet transform and traditional compression as well as reconstruction algorithm.Finally, we use matlab2010as a simulation tool. Compared with new method with the original method, the simulation results show that the method is feasible, reconstruction effect can meet the requirements of image processing. At the same time, in order to verify its paracticality, we debug programme in VC++6.0. At last, we get the desired results. At the same time, this method have hardware and software corresponding application environment.