Research On The Key Technologies Of License Plate Recognition In Mobile Terminal

As the fact that the average possession of vehicles is increasing, the phenomenon of traffic jam and traffic accident happen frequently. Intelligent Traffic Management(ITM) is an important research topic in the field of Intelligent Transportation System(ITS). Obviously, License Plate Recognition(LPR) is a key technology of ITM. There are many applications in modern traffic system such as EODS, Vehicle electronic toll collection system, Parking Administrative System. However, the traditional LPR with the fixed camera, the inflexible capture process and the single application scenarios is settled and unmovable. So it has certain limitations, such as not immediately register of illegal vehicles for a traffic management department. The key technology of LPR in mobile terminals remains in primary stage and the system performance based on this key technology is weak. Therefore, the LPR in mobile terminals has a value of further research.This thesis studies the key technologies of LPR which is suitable for mobile terminals based on image processing technology. The main research works are summarized as follows:1. A new method that achieved quickly and effective recognition is based on the difference of the color channel. The license plate area is highlighted rule of the color channel. And the final license plate location image is from the image of the connected area which is extracted through a variety of features selection.2. An improved method is put forward based on Radon transform to optimize the accuracy of skew emendation. And a method of vertical projection with character feature of width is used to segment characters.3. Aiming at the shortcomings of the traditional character recognition method, two methods about recognizing character are mentioned. One is an improved method based on multi-stage classifier. Another is based on SVM taking characters of coarse mesh, projection, invariant moment as the input. Meanwhile, the confusing character is second recognized after primary distinguish. On the basis of comparative analysis, the letter is recognized by the first method, and the second method is used to recognize the Chinese character and the mix of letter and Arabic numerals.4. As the key technology of license plate recognition is utilized in mobile terminal, a system based on Android is developed to query and record vehicle information intelligently and achieve the management for traffic management department in time.