摘要
本文在车牌定位的基础上,利用迭代法进行车牌图片二值化,然后利用基于连通区域标记的方法对图片进行目标区域标记,结合车牌先验知识对车牌字符进行切分。提取字符的13特征,利用BP神经网络进行字符识别。实验结果表明,本文采用的字符分割和识别算法效果良好。
On the basis of the license plate location, the paper focuses on binarizing the plate image by using iterative method, then carrying out region signature of the image by using connected component labeling algorithm, segmenting license plate character with the help of former knowledge; Extract thirteen features of the character, recognize the character using BP neural network. The results show that character segmentation and recognition algorithm in this thesis has obtained good results.
出处
《仪器仪表用户》
2009年第1期17-19,共3页
Instrumentation
关键词
二值化
字符分割
神经网络
binary
character segmentation
neural network
