摘要
介绍了基于三层BP神经网络的字符识别系统的实现。在VisualC++6.0开发环境下,采用BP神经网络开发的字符识别演示软件对于一定规范的字符能够正确检测识别,算法简单,易于实现。最后,探讨了BP神经网络算法与字符识别的优缺点。
This article introduces the design of a character recognition System used by three layer back-propagation neural networks. Under Visual C＋＋ 6.0, a character recognition software is developed using the algorithms mentioned above. Experiments show that the software is capable of detecting and recognising the characters which has Certain Norm. Above all, the arithmetic is very easy and is prone to realize. In the end, this paper probes into the advantages and disadvantages of the back-propagation arithmetic and the character recognition.
出处
《现代计算机》
2006年第10期93-95,共3页
Modern Computer
关键词
BP神经面络
字符识别
VC
Back-Propagation Neural Networks
Character Recognition
VC
