摘要
采用改进的BP神经网络,对汽车图像中车牌定位的方法进行了研究。选取训练样本图像,以人工定位后的图像为导师信号,将图像的灰度序列直接采用改进的BP算法进行训练,以期提高定位的精度和速度。试验结果表明,用改进的BP网络对灰度图像直接感知,在定位精度和速度方面取得较好的效果。
Does some research on detecting the license Number plate base on improved BP neural network. Some gray data images are chosen as swatches and some Images detected by manual are chosen. Improved BP algorithm is used to train the network and the precision and rate of management are advanced. The results show that this method is effective.
出处
《苏州大学学报(工科版)》
CAS
2004年第6期5-8,共4页
Journal of Soochow University Engineering Science Edition (Bimonthly)
基金
江苏省自然科学基金项目(bk99011)。
关键词
神经网络
车牌
BP算法
neural network
mlmber plate
BP algorithm