基于神经网络的车牌字符识别算法实验及程序校验

Vehicle Plate Recognition by Neural Net Work Based on Delphi

对算法简便、识别率高的BP网络和自适应谐振ART构成的自组织网络两种字符识别进行了综合实验与总结,结合程序论述了两种字符神经网络识别的原理并给出实验比较结果.

In this paper, two approaches are presented to recognize vehicle characters based on BP neural network and adaptive neural network constructed by ART,With codes examples,it describes two recognition principles and presents experimental results.