摘要
该文采用BP人工神经网络模型 ,研究汽车图像中车牌定位的神经网络方法。选取训练样本图像 ,以人工定位后的图像为导师信号 ,将图像的灰度序列采用BP算法进行训练 ,获得有效的网络权值。试验结果表明 。
This paper does some research on detecting the license number plate based on neural network.Some gray data images are chosen as swatches and some images detected by manual are chosen as tutors.BP algorithm is used to train the network and appropriate weigths are obtained.The results show that this method is effective.
出处
《南京理工大学学报》
EI
CAS
CSCD
北大核心
2002年第S1期104-108,共5页
Journal of Nanjing University of Science and Technology
关键词
神经网络
模式识别
车牌
neural network
pattern recognition
number plate
