摘要
介绍了一种BP神经网络数字识别方法,分析了从特征提取、网络模型建立、训练到识别结果的整个过程.由于运用特征提取,加快了网络的训练速度,也使网络更加健壮.通过实验,证明数字识别效果良好.
this poper introduces a method of digit recognition With BP neural network. ItShows the whole process of feature extraction, condruction of networks medel, trainingand the result of recognition. For using the feature extraction, the training is speedingup, the networks is stronger too. The result of digit recognition is satisfactory.
出处
《西安工业学院学报》
1995年第1期72-76,共5页
Journal of Xi'an Institute of Technology
关键词
数字识别
神经网络
BP神经网络
网络模型
recognition digital recognition feature extraction neural networks