Research On Character Recognition Of License Plate Recognition System Based On Neural Networks

The auto-realization of vehicle license character was mainly concerned in this paper.As for the part of the auto-recognition of the license character, a method based on Neural Network was presented. At last, we recognize the character of the license based on the method of neural network.In this article, neural pattern recognition is considered as the recognition method and vehicles' license plates is taken as recognition object, which is to improve the recognition rate of the character recognition of vehicle's license plates under noising circumstance. I have finished followed work hi article:1. Choose the feature of rough grid as the feature of character recognition, and improve the feature of rough grid. Optimize BP learning algorithm, then study single neural net classifier based on Optimized BP learning algorithm which takes rough grid as input. And use the single classifier on the figure, letter and province character recognition.2. To advance the recognition rate of the character recognition system of vehicle's license plates, we analysis the result of test, and put forward a series of measure. For example, design a careful neural net classifier, distill detail feature of character which is analogical, found stroke conglutination stylebook for some characters that have more stroke, and so on.3.Implement the correspondence arithmetic In Visual C++6. 0 based on the theory research and construct a software of vehicles' license plates recognition system.The research in this article shows that using rough grid feature for BP neural net classifier to recognize character can efficiently improve the robust and recognition rate of the recognition system. The method mentioned in the article has a certain reference value for the design of common the vehicle' license plates recognition system with random noise.