摘要
基于粗集(RS)理论和神经网络的车辆牌照识别系统由车牌定位和字符识别两部分组成。用主量分析(PCA)方法进行特征抽取,RS方法对特征信息进行预处理,约简后的特征作为网络输入,构成车牌字符识别网络。以“A”字母为例,提取字符特征后,并用RS化简神经网络的训练样本数据集,以实现车牌字符识别。
Recognition system of license plate based on rough set and neural network is consists of 2 parts for orientation and character recognition of license plate. Feature information is gathered with PCA, and it is preprocessed with RS way. Reduced feature is used as input of neural network, recognition network is composed with it. Take “A”as an example, after gathered feature information, learning sample data set of neural network is reduced by this method of rough
出处
《兵工自动化》
2004年第5期64-64,共1页
Ordnance Industry Automation
