摘要
通过对BP神经网络学习和数字字符识别问题研究,提出了一种基于改进的神经网络方法解决数字字符识别问题。试验利用Matlab7.0中的人工神经网络工具箱以及纯数字样本进行网络的学习训练。测试结果表明,该算法与传统BP算法相比,具有结构合理、收敛速度快的特点,能够很好地满足数字识别需求,达到了预期设计目的。
The paper studies the BP neural network and researches the recognition of the number characters,give a kind of method based on reformative neural network to solve the issue of number character recognition.The test uses the artificial neural network toolbox of Matlab7.0 and pure digital samples to train the network.Compare with traditional BP algorithm,the result reveals that this algorithm has structurerationalization and rapid constringency velocity,it can satisfy the requirement of character recognition,gain the prospective ends.
出处
《电脑编程技巧与维护》
2011年第12期90-91,115,共3页
Computer Programming Skills & Maintenance
