摘要
为了提高车牌字符识别率,将支持向量机SVM方法用于车牌字符的识别,算法首先采用Gabor变换和整体结构特征提取的方法提取车牌字符图像的特征参数,然后采用提取的特征训练SVM分类器,再应用SVM分类器分类和判别车牌字符。实验表明这种方法具有良好的车牌识别效果,较强的鲁棒性,有较大的应用价值。
An algorithm based on Gaber filter and Support Vector Machine（SVM） was proposed for vehicle license plate recognition. First, the features of plate characters Were detected by Gaber filter and global structural feature extraction. Then the features were used to train the SVM classifier. Finally, the plate characters were classified by the SVM ma- chines. Using the algorithm, a high recognition rate can be reached. Experimental results showed that the algorithm is feasible, robust and applicable.
出处
《四川大学学报（工程科学版）》
EI
CAS
CSCD
北大核心
2005年第5期130-134,138,共6页
Journal of Sichuan University (Engineering Science Edition)
关键词
支持向量机
车牌字符识别
特征提取
GABOR变换
support vector machines
license plate recognition
feature extraction
Gabor filter
