Abstract: Aiming at the problems of the character recognition for degraded license plates, a novel recognition method for Chinese character of license plate based on wavelet transform and fractal dimension is proposed in this paper. The image pre-processing and wavelet transform technologies are applied to the character images. An shifting differential box-counting approach(SDBC) is used to compute fractal box-counting of the images and gain the feature vector. Support Vector Machine(SVM) is adopted to design the classifier for recognizing the chinese characters . Experimental results show that the proposed method is robust in dealing with blurred character images, and it can improve the recognition rate of Chinese characters.

Key words: Chinese character recognition, feature extraction, wavelet transform, fractal dimension, Support Vector Machine(SVM)

摘要： 提出一种基于小波变换与分形维数的车牌汉字识别方法。对字符图像进行预处理和小波变换，应用改进的微分盒维法计算图像分形盒维值，并构造特征向量，利用支持向量机分类器对字符进行分类与识别。实验结果表明，该方法对模糊字符的识别具有鲁棒性，可提高汉字识别率。

关键词: 汉字识别, 特征提取, 小波变换, 分形维数, 支持向量机

CLC Number: 

• TP18