摘要
车标识别技术作为智能交通系统中的一项关键技术,对完善未来道路交通系统有着重大的意义。运用深度卷积神经网络对车标的识别进行研究,根据车标的特征,在经典的LeNet-5网络基础上对其进行细化研究,给出基于改进后的LeNet-5网络车标识别模型。为了验证基于深度卷积神经网络车标识别方法的可行性和有效性,采用深度学习框架Caffe,对改进的方案进行仿真分析。实验结果表明,改进的车标识别模型在外界环境的作用下依然具有较高的识别率,在现行的环境下更加适用于智能交通的发展需要。
In this paper,the deep convolutional neural network is used to study the identification of the vehicle standard.According to the characteristics of the vehicle standard,the refined LeNet-5 network is used to refine it,and the improved LeNet-5 network vehicle identification model is given.In order to verify the feasibility and effectiveness of the deep convolutional neural network vehicle identification method,the deep learning framework caffe is used to simulate and improve the improved scheme.
出处
《工业控制计算机》
2018年第12期36-38,共3页
Industrial Control Computer
