摘要
车牌识别技术是智能交通系统中的一项核心技术,由车牌定位、字符分割与字符识别三个部分组成。字符识别算法是车牌识别技术中的关键环节。然而,传统的基于SVM车牌字符识别算法针对多分类问题所需的分类器数量太多,导致训练时间太长且误差很大。因此,文中对SVM车牌字符识别问题构造了基于优化纠错输出编码的多分类,分类器数量大幅减少,从而节约了训练时间。实验结果表明文中提出基于多分类SVM的车牌识别算法识别率高,满足实际要求,而且训练时间短,鲁棒性良好。
License plate recognition technology is a core technology of intelligent transportation system.The license plate recognition system can be divided into three parts of the license plate location,character segmentation and character recognition. Character recognition algorithm is the key algorithm of license plate recognition algorithm.However,the traditional classifier based on SVM for License Plate Recognition required so much classifiers that needs long training time and leads to a large error. Therefore,this paper proposes a new algorithm based on an optimized multi-classification ECOC classifiers for the issue of SVM license plate character recognition,which significantly reducing the number of classifiers and saving training time.Experimental result shows that the proposed license plate recognition algorithm based on multi-classification SVM has a high recognition rate and a good robustness,meeting the practical requirements,and the training time is much shorter.
出处
《物流工程与管理》
2016年第5期260-263,共4页
Logistics Engineering and Management
