摘要
车牌识别技术作为智能交通系统的核心,起着举足轻重的作用.通过将独立成分分析和小波矩方法引入到车牌字符识别研究中来,提出了一种对车牌字符按照汉字组、数字与字母组进行分组,然后对不同组提取不同的特征进行识别的新方案.实验结果证明,所提方案可以达到很高的识别率.
Plate characters recognition is the kernel part of LRP. The authors proposed a new method based on ICA and wavelet moments, firstly, divide the plate characters into two classes; secondly extract different features for the two classes; at last recognize them with BP network classifier. Experiments show this method has a good performance.
出处
《西南师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2010年第2期159-163,共5页
Journal of Southwest China Normal University(Natural Science Edition)
关键词
车牌字符识别
独立成分分析
小波矩
plate characters recognition
independent component analysis
wavelet moments
