摘要
汽车车型自动识别在现代交通监控和管理中有着广阔的应用前景,本文探讨了用于汽车车型识别的BP神经网络分类器的设计问题。在简述车型图象预处理和特征提取的基础上,着重阐述学习算法的选取问题,提出一种分阶段学习方法。仿真结果表明这是一种有效的学习算法。
The automatic recog nition of automobile type has a promising and practical future in the traffic control and the desig ning problem of the BP neural network classifier used in recog nizing the automobile types.On the basis of the pretreatment of the imag es of the automobile type and character abstraction,it mainly set forth how to choose the learning alg orithm and put forward a phase learning method.The simulation result in-dicates that it is an effective learning alg orithm.
出处
《工业控制计算机》
2003年第1期19-21,共3页
Industrial Control Computer