摘要
复杂背景下车牌字符的分割包括:预处理,确定车牌字符排列格式;聚类分析字符分割,即通过计算字符宽度、合并相邻数域等步骤处理车牌字符;车牌字符的特征提取含基于Sanger算法的主分量分析遗传神经网络、字符图像的重建和字符特征主分量的可信度度量。实验验证该算法具有准确率高和鲁棒性强的特点。
The segmentation of license-plate character includes preprocessing and definition of character arrangement. Cluster-based of character segmentation involves computing character＇s width, combination of adjacent domain and so on. Feature extraction is divided into three steps of principal components analysis combined with genetic-neural network based on Sanger algorithm, reconstruction of character and reliability on character principal components. The experimental results show that the algorithm has high accuracy and robustness.
出处
《兵工自动化》
2005年第4期69-71,共3页
Ordnance Industry Automation
基金
湖南省教育厅自科基金资助课题(01C078)
关键词
车牌字符
字符分割
聚类分析
特征提取
主分量分析
遗传神经网络
License-plate character
Character segmentation
Cluster-based analysis
Feature extraction
Principal comnonents analysis
Genetic-neural network
