针对车牌字符识别的特点, 提出了基于二维小波包的车牌字符特征提取方法, 将车牌字符图像进行二维小波包变换, 搜索最优小波包基, 提出了寻找最优小波包基的准则, 并给出了具体算法; 将获得的最优小波包基作为车牌字符的特征, 采用神经网络进行车牌字符分类. 测试结果表明, 与传统方法相比, 文中方法得到了更好的结果. Aiming at the characteristics of license plate character recognition, a method of feature extraction of vehicle license plate characters based on two-dimensional wavelet packet is proposed. The license plate character image is transformed by two-dimensional wavelet packet to search the optimal wavelet packet basis. And gives the specific algorithm.With the optimal wavelet packet basis as the feature of the license plate characters, the license plate character classification is carried out by using the neural network.The test results show that the proposed method has better results than the traditional method.