Abstract: A kind of character recognition method of the vehicle license plate based on the wavelet packet and Zemike moments is presented in this paper. Firstly, the digital character is processed by mathematics morphology for its special typeface structure. Then, the character is analyzed by wavelet packet for three layers to get the coefficients in nodes of the third layer. In addition, the nodes of the second layer are reconstructed and the Zemike moments of the reconstructed images are computed individually. At last, the coefficients of the third wavelet packet and the Zemike moments make up the feature space, which is processed by reducing the dimension. The digits of the vehicle license plate are recognized by BP neural network. The experimental results demonstrate the efficiency of the proposed approach.

Key words: character recognition, wavelet packet, zernike moments, neural networks

摘要： 本文提出了一种基于小波包和 矩特征提取的车牌字符识别方法。首先针对数字字符特殊的字体结构，采用了数学形态学方法进行滤波处理。预处理后，对待识别字符进行三层小波包分解，获取第三层各节点小波包系数；同时对小波包分解后的第二层各节点进行重构，并分别计算重构后所得图像的 矩；最后对获得的小波包系数和重构后所得图像的 矩所组成的特征空间进行降维处理,并利用BP神经网络对车牌中的数字进行识别。实验结果表明，该方法效果良好。

关键词: 字符识别, 小波包, zernike 矩, 神经网络