The Research Of License Plate Recognition System Based On Neural Network

The License Plate Recognition(LPR) system is an important part of Intelligent Traffic System(ITS), which is very valuable in practical applications such as public safety . traffic management and military department. Based on the prior work, LPR system software is improved, we made further research focusing on theories and technologies of license plate characters segmentation and recognition, character feature extraction and recognition is our concern.1. The Principal Component Analysis(PCA) method based on the Singular Value Decomposition(SVD) is applied in the feature extraction of license plate characters, the principal component is reduced by Rough Set(RS), the fewer features are taken as input of neural network, thus not only the complexity of network structure is decreased, but the rate of learning and the recognition rate of characters are enhanced.2. In the research of identification methods, the advantages and disadvantages of model match and neural network are compared and analyzed, we combined these two methods and designed some classifiers using a multilevel-multiclassification scheme, which can classify numbers, letters, alphanumeric characters and Chinese characters. The result shows that it is more effectively in similar character recognition.