| With the amount of vehicles in China has been increasing in recent years, the problem of counterfeit vehicle license plate has come into being and even become more and more serious. The main reason is that this sort of the illegal vehicle can not be effectively identified. A prediction system, used to identify whether the vehicle used fake license, is presented in this paper. This method take vehicle picture with video monitoring equipment, and combine with the image processing algorithm to automatically identify the model and license number of vehicle. Then, comparison between identifying information and database information is used to identify whether the vehicle used fake license. In recent years, crimes of counterfeit vehicle increase continuously with diversified tricks. This kind of vehicles have the same vehicle model, even the same color and the same license number with the registered one, we simply call them SMCL vehicles. Such kind of vehicle can not be identified by image processing technology. Prediction system of counterfeit vehicle license plate based on the shortest path algorithm which used to identify whether the SMCL vehicle used fake license, is presented in this paper.A rapid license plate localization algorithm which is the combination of color image segmentation and plate texture features is presented in this paper. This method can accurately locate the position of license plate in the picture quickly. Then used the template matching method to complete the initial character recognition, need extension of the similar code and single character. The method based on the theory of Fuzzy Pattern Recognition is divided into three parts. Firstly, use Hough transformation to extract the feature points of vehicles, and use the ratio between two absolute distance of adjacent feature points as the characteristic values of vehicles; secondly, use Fuzzy C-mean Classification to handle feature data of 75 car model, then establish a degree of membership matrix as the sample space; thirdly, consider the classification algorithm based on fuzzy approach degree and the credibility of the vehicle feature to propose a weighted close- degree recognition algorithm. This recognition method has a good effect. This paper identifies the fake license plate according to the rule of "a car can't be in two different locations at the same time". The flow of this method is as followed: firstly, solving the shortest path between two monitoring position by improved ant colony algorithm. According to the ant colony algorithm possibly has slow convergence speed and tends to be trapped in local optima, a degree of polymerization is presented to judge the situation of ant colony, to increase or decrease the attraction of pheromone for ant colony. Meanwhile, oversize pheromone should be avoided by setting the maximum and minimum value of pheromone for every path. This paper proposes an improved elliptic defined region search algorithm. It is greatly narrowed the search area, to avoid the unnecessary path of the ant colony. Compared to other similar algorithms, this improved ant colony algorithm which is presented in this paper has higher identification accuracy; the recognition speed of this method has greatly improved.The shortest path is one of the important factors for identifying SMCL. The shortest path can be calculated as the quantitative factor. Then combine with the time credibility degree and the traffic unobstructed degree in this period of time, whether the vehicle use fake license plate can be decided. In the system, a T-S fuzzy neural network is used to comprehensively evaluate the forewarning grade of the SMCL vehicles. The vehicle with extremely suspicious grade can be automatically indicated. Through the experimental simulation, the error of the neural network is in allowing scope. So it can be used in the early warning system for the counterfeit vehicle. It provides a new idea for the identification of counterfeit vehicle license plate. |
PDF Full Text Request